mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-10-27 11:20:42 +01:00
0cb3fc7db591ba23e21fbbc70042bc44ec7d6f20
1903
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 Devin FoleyandPaperclip
|
0cb3fc7db5 |
fix(runner): retain ACPX transport failure details (#17837)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Paperclip Runner reports provider failures to the execution and recovery services. > - Those services read the terminal event's structured `error` field. > - The ACPX transport-failure path wrote its code and message only at the top level. > - The session facade therefore replaced that cause with `error: null` and the generic message "Provider turn failed". > - This pull request uses the existing structured error contract while retaining the old fields. > - Recovery can retain the actual redacted cause without reopening the failed provider or assuming a retry is safe. ## Linked Issues or Issue Description **What happened?** A native ACPX run can lose its failure details between runnerd and the session facade. The transport-failure producer emits a failed turn with top-level `code` and `message`. The facade reads `error`. The server receives a generic provider failure and cannot use the original cause to explain or classify recovery. **Expected behavior** Retain the transport code and redacted cause in the structured terminal error. Keep failed sessions closed, provider termination unproven, and automatic replay disabled. **Steps to reproduce** 1. Open a native ACPX session and start a turn. 2. Send malformed activity from the sidecar after it acknowledges the turn. 3. Poll the native provider events. 4. Observe that `turn.failed.error` is absent even though the producer has a transport code and cause. Refs: #16366, #17543. Those changes cover legacy process-session diagnostics and session-ensure errors. This change repairs the native ACPX polling terminal contract. ## What Changed - Add the existing transport code and redacted message to the terminal event's structured `error` field. - Mark that error as non-recoverable. Keep the previous top-level fields for existing readers. - Add a real sidecar fault fixture and a native-provider regression. It checks one failed terminal, retained error details, a closed session, and no provider reopening. - Document the failure contract and verification scope in execution semantics and the recovery map. ## Verification - Red: the new fault test failed because the structured transport error code was null on the original code. - Green: all 26 native-provider integration tests pass with the fix. - Three session-facade terminal tests and three native-runtime provider-failure classification tests pass. - `pnpm -r typecheck` and `pnpm build` pass. - Hosted test partitions, native Runner checks, typecheck, build, browser shards and canary passed. Two final aggregate jobs received no runner for several minutes. The first CI attempt was cancelled after saving its status, and only those two cancelled jobs were resubmitted. Both aggregate jobs passed on attempt 2. The current head has 55 successful checks, four expected skips and successful Snyk status. - `git diff --check`, `node scripts/check-module-boundaries.mjs`, and `node scripts/check-maps.mjs` pass. - The full local Vitest suite is not run for this patch. The local filesystem fell below 500 MiB free during verification. The preceding full-suite attempt crashed embedded PostgreSQL; focused checks and CI are used without claiming that attempt passed. - Live incident: verified the exact predecessor sandbox was destroyed and workspace recovery completed. An audited user retry through the normal task-retry service started one new native run. That attempt preserved the original question but failed before provider startup with a missing workspace-sync descriptor binding. The startup race is fixed separately in #17838. A later audited retry completed successfully, saved its useful answer and restored its workspace. The agent API handoff failed; an audited operator repair through the normal issue service completed the required user assignment in Review without repeating the comment or provider work. This verifies recovery admission on the existing deployment; the new Runner change is not deployed yet. - Evidence limit: the historical terminal had `error: null` and its raw sidecar frame was not retained. Fault injection proves this producer defect and the same loss of diagnostics. It does not establish which transport fault caused the historical turn to end. ## Risks - Recovery now receives a specific transport failure code. The error remains non-recoverable and `providerTerminalObserved` remains false. Stop proof, effect reconciliation, workspace preservation, retry limits, and normal execution authority still govern any successor. - This change preserves diagnostics. It does not make a malformed or lost provider connection healthy. - No schema migration, credentials, provider selection, or model settings change. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, source inspection, tool use, and code execution. The runtime does not expose the exact model ID or context window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
commitperclip[bot]
|
2d72f2c4c0 |
fix: Serialize ACPX session record saves per record to stop overlapping renames (EPERM on Windows) (#16925)
> [!NOTE] > Opened by Prospector's automated issue-fix pipeline. A maintainer reviews and decides; nothing here merges on its own. ## Thinking Path - Issue #15346 reports a defect in paperclipai/paperclip. - Root cause, as the fixing agent read it: ACPX saves one session record from several writers (live checkpoint timer, turn end, finalize, close) with no coordination; each save renames a temp file onto the same record file, and on Windows overlapping renames onto one file fail with EPERM. All of these saves go through Paperclip's session-store wrapper, which forwarded them concurrently. - This pull request: Serialize ACPX session record saves per record to stop overlapping renames (EPERM on Windows) ## What Changed - `packages/adapter-utils/src/acpx-engine/ephemeral-session-environment.ts`: Every ACPX save in Paperclip passes through this wrapper, so it now puts all saves of one record id in a single queue (a promise chain shared by all wrappers in the process). Saves of one record never overlap and keep their order; a failed save still returns its own error and does not block the next one; saves of different records still run in parallel. ## Verification - Reproducing test(s), added in this PR: `packages/adapter-utils/src/acpx-engine/ephemeral-session-environment.test.ts` - They fail on the base commit and pass with the fix, each twice in a fresh sandbox. - 3 independent attempts wrote 3 reproductions; this fix passes every one of them. - The repository's compile command passes with the fix applied. - 3 existing test file(s) near the change pass. - The full test suite passes except for the 246 file(s) that already fail on the base commit. ## Risks - Written by an automated pipeline, not a person: review it as you would an outside contribution. - Risk tier of the touched paths: 2. - The scope reviewer found it also changes, beyond the report: save() called when no earlier save of that record is pending - The scope reviewer found it also changes, beyond the report: persistedStore.save throwing synchronously - The scope reviewer found it also changes, beyond the report: Records with missing/undefined acpxRecordId - The scope reviewer found it also changes, beyond the report: save() for different record ids ## Model Used - opus, sonnet (Anthropic Claude, via Claude Code), writing and checking the change in an isolated sandbox. ## Linked Issues Fixes #15346 Other pull requests that name #15346: - #15370 (closed): fix(acpx): serialize session file saves and clean failed temporary writes ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [ ] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with 'Fixes: #' / 'Closes #' / 'Refs #' OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub '#NNN' / 'github.com/paperclipai/paperclip' URLs) - [ ] My branch name describes the change (e.g. 'docs/...', 'fix/...') and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [ ] I will address all Greptile and reviewer comments before requesting merge <!-- prospector:issue-fix v1 issue=15346 base=1c07b5903b1b report=4fe3fefc8864e00f --> Co-authored-by: commitperclip[bot] <284812584+commitperclip[bot]@users.noreply.github.com> |
||
|
Andrii Furmanets
|
61ecc93033 |
fix: count a recovered shutdown interrupt as recovered (#15031)
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work
> - The dashboard success rate counts recovered runs as successes
> - A run is recovered when a later retry in its chain succeeds
> - The summary puts `failed` and `timed_out` into recovered
> - An `interrupted` run, including `server_shutdown_interrupted`, falls
through to other
> - The success rate stays down after the retry succeeds
> - This pull request counts a recovered interrupt in recovered, not in
other
> - The benefit is that a shutdown that was retried successfully no
longer looks like a failed day
## Linked Issues or Issue Description
No open issue covers this. I searched open and closed issues and pull
requests.
**What happened?**
`dashboardService.summary` puts only `failed` and `timed_out` into the
recovered bucket. An `interrupted` run, including
`server_shutdown_interrupted`, goes to other even when `recovered` is
set. The success rate uses `succeeded + recovered`, so the day stays low
after a successful retry.
**Expected behavior**
When the recovered flag is set, count `interrupted` in recovered. An
interrupt with no successful retry stays in other.
**Steps to reproduce**
1. Record an interrupted run with error code
`server_shutdown_interrupted`.
2. Record a succeeded retry of that run.
3. Open the dashboard summary for that day.
4. Observe that the interrupt is in other and the success rate does not
include it.
**Paperclip version or commit**
master at
|
||
|
mdc2122andmdc2122
|
8153061d71 |
fix(pi-local): strip inherited server PAPERCLIP_* env from pi_local processes (#14503)
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work
> - The `pi_local` adapter spawns the Pi CLI as a child process for
agent runs, environment tests, and model discovery
> - Each of these three paths builds the child env with `{
...process.env, ...env }`, so the child inherits the full server
environment
> - The server environment can hold server-only `PAPERCLIP_*` secrets
and config paths, such as `PAPERCLIP_AGENT_JWT_SECRET`,
`PAPERCLIP_TOOL_ACTION_SIGNING_SECRET`, and
`PAPERCLIP_SECRETS_MASTER_KEY_FILE`
> - `@paperclipai/adapter-utils` already has
`sanitizeInheritedPaperclipEnv`, and the shared `runChildProcess` path
uses it for its base env. `pi_local` passes its pre-merged env as
`opts.env`, which `runChildProcess` spreads over that sanitized base, so
the inherited server values come back
> - This pull request routes the three `pi_local` merges through
`sanitizeInheritedPaperclipEnv`
> - The benefit is that `pi_local` agent processes no longer receive
server secrets. This is a security hardening change.
## Linked Issues or Issue Description
No issue exists. This is a bug report in the in-PR format.
**What happened?**
A `pi_local` agent process receives the inherited `PAPERCLIP_*`
variables of the Paperclip server process. An env dump from inside a
`pi_local` run showed `PAPERCLIP_AGENT_JWT_SECRET`,
`PAPERCLIP_TOOL_ACTION_SIGNING_SECRET`,
`PAPERCLIP_SECRETS_MASTER_KEY_FILE`, and server config path variables.
The agent process can read these values.
**Expected behavior**
The `pi_local` child env drops inherited server `PAPERCLIP_*` variables,
the same as the shared `runChildProcess` path. Only
`PAPERCLIP_RUNTIME_API_URL`, `PAPERCLIP_LISTEN_HOST`, and
`PAPERCLIP_LISTEN_PORT` stay inherited. The run-scoped variables that
the adapter sets (for example `PAPERCLIP_AGENT_ID` and
`PAPERCLIP_API_KEY`) and the adapter-configured env still reach the
child.
**Steps to reproduce**
1. Start the Paperclip server with `PAPERCLIP_AGENT_JWT_SECRET` set in
its environment.
2. Run a `pi_local` agent whose command writes `process.env` to a file.
3. Read the file. `PAPERCLIP_AGENT_JWT_SECRET` is present.
**Agent adapter(s) involved**
`pi_local` only.
**Additional context**
Other CLI adapters appear to merge `process.env` into the child env in a
similar way (for example in some `execute.ts` and `test.ts` files). This
pull request does not change them. It keeps the scope to `pi_local`.
Related: #3030 and #12870 also touch inherited agent env. This pull
request is independent of #14479.
## What Changed
- `packages/adapters/pi-local/src/server/execute.ts`: the run env and
the runtime env that the adapter logs now start from
`sanitizeInheritedPaperclipEnv(process.env)` instead of `process.env`.
- `packages/adapters/pi-local/src/server/test.ts`: the environment test
env now starts from `sanitizeInheritedPaperclipEnv(process.env)`.
- `packages/adapters/pi-local/src/server/models.ts`: the model discovery
env (`pi --list-models`) now starts from
`sanitizeInheritedPaperclipEnv(process.env)`.
- `server/src/__tests__/pi-local-execute.test.ts`: add a regression
test. The test sets three server secrets on `process.env`, runs
`execute` with a fake Pi command that dumps its env, and checks that the
secrets are absent. It also checks that the adapter-configured env and
the run-scoped `PAPERCLIP_AGENT_ID` and `PAPERCLIP_API_KEY` are present.
## Verification
- `cd server && npx vitest run src/__tests__/pi-local-execute.test.ts
src/__tests__/pi-local-adapter-environment.test.ts
src/__tests__/pi-local-skill-sync.test.ts`: 7 tests pass.
- The new test fails without the source change (the secret assertion
fails) and passes with it.
- `npx vitest run packages/adapters/pi-local`: 42 tests pass.
- `pnpm --filter @paperclipai/adapter-pi-local typecheck`: pass.
- `cd server && tsc --noEmit` (after `pnpm --filter
@paperclipai/plugin-sdk ensure-build-deps`): pass.
- Follow-up coverage observes real child environments from direct model
discovery, agent execution, and `testEnvironment`. Each path omits
inherited secrets and retains the permitted runtime/listen variables and
configured env. Removing either the model-discovery or environment-probe
sanitizer independently makes the regression fail. All 49 scoped tests
and both direct typechecks pass.
## Risks
- Low risk. A `pi_local` setup that relied on an inherited server
`PAPERCLIP_*` variable (other than the three kept keys) no longer gets
it. Such a value can still be set in the adapter `env` config. This is
the same behavior as the shared `runChildProcess` path.
> For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and
discuss it in `#dev` before opening the PR. Feature PRs that overlap
with planned core work may need to be redirected — check the roadmap
first. See `CONTRIBUTING.md`.
## Model Used
- OpenAI GPT-6.1 Sol, exact runtime model ID
`gpt6sol-cliproxy/gpt-6.1-sol`, with tool use and code execution.
Context-window and thinking-level metadata are not exposed. The prior
model and human-review disclosure was not substantiated by available
runtime metadata.
## Checklist
- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge
---------
Co-authored-by: mdc2122 <259970464+mdc2122@users.noreply.github.com>
|
||
|
DottaandPaperclip
|
dcfc2acfa8 |
Preserve complete bounded OpenCode MCP tool results (#17825)
<!-- entire-trail-link-start --> https://entire.com/gh/paperclipai/-/trails/34/changes/paperclip/34 <!-- entire-trail-link-end --> ## Thinking Path > - Paperclip manages AI agents and their governed tools. > - Native OpenCode receives tool results through a Runner-owned MCP bridge. > - The bridge cut serialized results at 64 KiB, even inside a JSON string. > - A tool could finish while its saved OpenCode output lost files and receipts. > - Native app results already use bounded results and saved-result pages. > - This change preserves complete JSON within a bounded output limit. > - Agents can read the result without repeating the completed action. ## Linked Issues or Issue Description **What happened?** The OpenCode MCP bridge sliced serialized JSON at 65,536 characters. A large file-list result became invalid JSON in OpenCode's saved output. The tool execution itself had completed. **Expected behavior** The bridge must deliver complete bounded JSON. A delivery limit must produce an explicit error, rather than a partial success value. A repeated call identity must not execute the action again. **Steps to reproduce** 1. Return a structured result with 27 file entries and duplicated text/data representations larger than 64 KiB. 2. Call the tool through the actual HTTP MCP bridge. 3. Parse the returned text as JSON. The old source fails at position 65,536. Related: #17740 changed assigned-tool catalog delivery. This PR repairs result delivery and does not change the tool catalog or authority. ## What Changed - Replace JSON slicing with complete serialization and a 1 MiB UTF-8 output limit. - Bound serialization traversal, including escaped strings and many small fields. - Keep the existing native 500 KB result/page and saved-artifact contracts. - Report oversized delivery with a warning that the action may already have completed. Keep call-identity replay on the original execution. - Add real HTTP regressions and document the transport limit. ## Verification - Red before fix: the large-result HTTP regression fails with an unterminated JSON string at position 65,536. - 91 focused MCP, OpenCode driver and continuation assertions pass on Node 24.22.1. They cover escaped UTF-8 pages, oversized delivery, authentication, cancellation and duplicate-call execution. - Workspace typecheck and build pass. - Actual native Rust → fixed MCP bridge → OpenCode 1.18.34 → `openrouter/deepseek/deepseek-v4.1-flash-1017` passed with a public synthetic 27-file result. One assigned read produced a 281,375-byte saved JSON file. The model used Bash/Python to parse all filenames and both unpredictable receipt fields. One finish completed the turn. - The native probe used a synthetic gateway and a mocked completion acknowledgment. It does not prove hosted deployment, board finalization or a particular upstream model vendor. - The full default test attempt is failed/incomplete. It was stopped after repeated embedded PostgreSQL bootstrap failures and skipped database suites. Its 15 warm-workspace cases pass unchanged on isolated PostgreSQL 18. A controlled default-database rerun fails before the test body when embedded PostgreSQL cannot start after five attempts. The test source is byte-identical to the base commit. The failed attempt remains recorded. - The final test-only commit checks a 500 KB escaped producer page inside the actual controller text envelope. The production bridge hash still matches the native proof. Its 91 focused tests and package TypeScript check pass. - Current-head verification (`73d18ca2`): 53 CI checks passed and two conditional Storybook checks skipped. Greptile is 5/5 on this exact head with no review threads. The failed/incomplete full local attempt and the controlled passing suites above remain separate evidence. ## Risks - Larger complete results use more bounded serialization memory than the old truncated output. - A bridge result above 1 MiB fails delivery. Native app results and saved-result pages are bounded at 500 KB before this bridge. The existing artifact protocol handles larger results. - A failed delivery cannot prove whether an upstream action took effect. The error forbids repeating it, and identical call identities retain the original execution. - This change does not alter model routing, admitted tools, credentials or trust policy. Hosted rollout and incident recovery remain separate verification steps. ## Model Used - OpenAI `gpt-6.1-sol`, with reasoning, tool use and code execution. The session did not expose an authoritative context-window size. Native acceptance used the exact DeepSeek model named above. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run the relevant local tests and they pass; the full local-suite environment limitation is documented above. - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
208a093d19 |
fix(chat): preserve resumed review waits and retry newer saved input (#17725)
<!-- entire-trail-link-start --> https://entire.com/gh/paperclipai/-/trails/36/changes/paperclip/36 <!-- entire-trail-link-end --> ## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - GitHub review agents publish findings and wait for an authorized reply or head change. > - Review tools could retain the requester after resume while completion lost that same wait proof. > - A yielded report could then queue its own continuation or repeat a missing-wait error. > - Ordinary Retry could also select an older failed turn while a newer saved request had no admission. > - This change shares the existing authority proofs and canonical admission owners. > - Current access, publication, execution and budget gates remain required. ## Linked Issues or Issue Description Refs #17318. Builds on the canonical saved-input retry owner in #17733. Related #16550 covers deferred legacy comments after ownership waits. The separate recovery repair in #17724 owns the cleanup-to-saved-input handoff; this PR does not add a second admission owner. A resumed review could publish an assessment, then turn its `response_wake` into an immediate `issue_status_changed` wake. A provider that omitted its continuation could instead repeat a completion error. A newer linked-human message saved during a budget hold could also remain unstarted because Retry selected the older failed source. ## What Changed - Resolve passive review waits through the accepted requester ledger used by GitHub tools. Require the latest reviewed head, exact company, task, agent, repository and processed publication under current access. - Fill omitted/null yielded continuation from that server proof. Permit an inherited normal tree resume only through its accepted parent identity. Permit an ordinary follow-up only after this run has a processed useful reply linked to its exact inbound message. Reject acknowledgements, partial assessments, pending or uncertain writes as that reply proof. - Recheck the same run/review/head/source proof during feedback, finalization and final status commit. Preserve explicit malformed-input rejection and Done policy. Add no monitor permission or publication grant. - Keep heartbeat and publication projections narrow. Preserve both published Copilot v62 identities and generate source-bound v63 from the pinned merged base. No v63 live qualification is claimed. - Certify the structured isolated-directory binding failure only at its pre-allocation boundary and only through canonical task Retry. Other workspace validation reasons and unknown effects remain denied. - Route Retry to eligible newer saved input through the existing canonical owner. Preserve the old result and skipped receipt. Require current principal, source, App, runtime fence, budget, native ownership and exact cleanup proof. - Recover absent historical App metadata only through an earlier trusted processed response receipt and an immutable non-null endpoint claim. Explicit malformed or conflicting metadata stays denied. - Revalidate execution through the exact claimed wake, coalesced task Retry and resolved user-continuation receipt after admission. A second normal Retry can recover the exact never-started authorization failure with unchanged saved input. Require empty execution evidence, exact diagnostics and no allocated or active executor. - Explain when a recorded result prevents exact replay. Ordinary board Retry directs a new task message to continue from saved work. Inherited requester identity alone never grants replay. - Keep accepted board-message preparation failures with the canonical task Retry owner. Certify the exact unchanged comment, original user wake, accepted same-user identity and absence of provider execution. Pin that source, recheck it at admission, and bind identity initialization to the coalesced intent and exact claimed successor receipt. Uncertifiable sources retain an actionable fresh-message refusal. - Return safe named authority categories for board Retry refusals. Update execution documentation and capability maps. Add no schema or polling loop. ## Verification - A fresh external review request exposed a separate isolated-directory preparation failure before allocation. Its first normal Retry saved an intent but admitted no successor. This follow-up certifies only that exact structured preparation boundary through the existing canonical Retry owner, with trusted failed settlement and no executor, lease, result or output. A real HTTP/queue regression reproduces the missing successor before the repair. Twenty-three current-source safety cases and the final 98-case Retry/HTTP group pass, including a fresh click after a denied intent, source revocation, accepted results and the older direct API remaining denied. The warning case uses actual configuration preparation and production event append; unknown warning shape or provider provenance stays denied. Full workspace typecheck, build, maps and diff checks pass. Renewed hosted normal Retry now succeeds on the final combined source: it preserves the exact saved request and linked requester, starts one native successor in a fresh attempt workspace, publishes the current-head 2/5 assessment and commits In review with passive response wake. - A later hosted normal Retry published a low-score assessment but lost its accepted source identity because the previous attempt failed before identity initialization. It repeated continuation errors, then submitted Done. This is failed acceptance, not a successful qualification. The initializer now retains the exact source only after execution-time failed-chat Retry authorization; generic Retry and Done policy are unchanged. A real queue/initializer regression fails with a null source on the previous code and passes with the newer saved source. Two real database completion cases verify that source yields the authoritative wait and commits In review. Renewed hosted normal Retry now retains the accepted source and exposes the authoritative current-review wait. It completed after two semantic feedback corrections (wrong contract revision, then wrong criterion IDs); the correct contract was present in emitted input. The third finish was accepted and committed In review, with no missing-wait loop or Done escalation. - Repeated board-message preparation failures now walk their canonical intent, failed dispatch and accepted parent identity, rechecking each contained predecessor. A production queue regression fails the second ordinary HTTP Retry before the repair and passes afterward. Changed source, contradictory execution evidence and cyclic receipts remain denied; traversal stops after twenty links. - New inherited omission cases failed before the repair. The final full PostgreSQL 18 wait/status suite passes **252 tests** under Node 24.22. It covers omitted/null input, malformed-first recovery, same-head resume, exact-source replies, changed heads, revoked access, pending/uncertain/mixed publications, final commit revalidation, no duplicate review and no self-wake. - **15 existing publication tests passed before the pinned merge** on a separate fresh database. They retain default partial-delivery and failure-fallback behavior. Context reads remain below the unchanged 16 KiB heartbeat and publication response limits with large saved output present. - Saved-input tests cover a real signed PR webhook, production accounting-budget refusal, normal policy recovery and canonical Retry. Historical producer-shaped tests prove the missing-App case red, then check trusted prior receipts and malformed, late, revoked or mismatched proof. A broader order-coupled test was reproduced on the prior source; its global fault injector now targets its own authorization query. The 43-case sequence passes without changing production behavior or its safety assertions. - New signed-ingress and real heartbeat execution tests reproduced the admitted-successor failure, then the next normal Retry failure. The final pinned merged source passes 249 full GitHub review/publication and Retry cases on a fresh database, including all 49 saved-input cases. A separate 74-case Retry/HTTP run also passes. Exact dispatch negatives cover source, company, agent, task, successor, receipt, pause, assignment, accepted success, controller and cleanup proof. Bootstrap contradictions remain denied. The canonical Retry suite passes 16 cases through its supported external PostgreSQL test connection. - The normal board HTTP route reproduced the unusable inherited-result refusal. It now returns actionable guidance without starting a successor. The actual UI Retry hook and banner tests pass 12 cases. - A production-shaped accepted board comment followed by a pre-provider qualification failure reproduced HTTP 409 before the repair. The final real HTTP/queue test certifies one canonical successor and preserves the accepted identity. It rejects changed source, actor, task, agent, result, controller, process, profile, allocation and receipt evidence. A post-acceptance source change keeps the same intent pending. Execution revalidation rejects removed or mismatched proof. The probe stops deliberately after production identity initialization; it does not claim provider execution. The final source passes the 75-case Retry group, the focused two-case board/inherited-policy group and 19 canonical Retry cases. Full workspace typecheck, build and map checks pass. - Full workspace `pnpm -r typecheck` and `pnpm build` pass. Map, module-boundary and diff checks pass. Full local `pnpm test:run` remains unverified. The 256-case explicit-native suite cannot initialize its packaged PostgreSQL because libzstd is absent; an existing command-scoped fallback did not resolve it. Current-head remote full-suite checks are required. Merged-source full GitHub+Retry (249), separate Retry/HTTP (74), full wait/status (252), profile/history and completion (75), and visible Retry hook/banner (12) checks pass. The fresh owned PostgreSQL 18 databases applied all 377 migrations. - Actual Rust Runner, OpenCode 1.18.34 and exact DeepSeek used unchanged production-generated synthetic contexts. Resume finished with an explicit server wait. The author-reply trial omitted continuation twice: its first report failed the existing empty-criteria check; the corrected second report still omitted continuation and received the exact server wait. It completed once with no missing-wait loop or Done escalation. Controller acknowledgments were synthetic; these probes do not establish hosted publication or board state. - Hosted combined testing proved automatic admission of the exact saved author reply, a useful GitHub answer and committed passive wait. It required two malformed-continuation rejections before a nullish fallback succeeded. A later normal Pause/Resume completed in one finish with explicit response_wake and no new review/comment. The admitted normal Retry then exposed the execution guard gap repaired here. Those remain checkpoint results. On the final combined source, renewed normal Retry completed actual native provider work and GitHub publication, committed the passive wait, released ownership and cleaned up its environment. A subsequent ordinary unmentioned author reply woke once, made no GitHub tool calls or new assessment, and completed its first finish with an explicit response wake. Both exact leases expired with successful cleanup and the queue was empty. However, final destination inspection found a separate automatic fallback overwrote the existing summary with a generic no-final-response message. External summary preservation therefore remains failed acceptance; the separate publication repair in #17726 owns that producer. Full combined publication preservation has not passed. These hosted runs include the pinned-base native additions; they do not qualify Copilot v63. - Current-head automated review is 5/5 with zero unresolved threads. All current-head CI gates passed (54 successful checks/contexts, four skipped, none pending). The earlier login-cleanup idle-work assertion also reproduced locally while its controlled baseline passed; its cause remains unknown, but the exact assertion passed in the fresh current-head CI shard. No timeout, assertion or production workaround was added. This PR is ready for review within its source, authority, native finalization, state, wake and cleanup scope. The independent automatic fallback publication regression is tracked by #17726; no complete combined workflow pass is claimed. ## Risks - The latest saved review defines the durable head boundary. Ingress must record a new head before admission. A reply receipt proves delivery only for its exact saved source; a shared GitHub comment ID alone cannot authorize this wait. - Historical App recovery cannot infer identity without direct proof or an earlier valid server response receipt. It never replays that response. - This does not infer review-score policy from instructions or override Done. No deployment, merge or repair of historical tasks is claimed by the PR. ## Model Used OpenAI `gpt-6.1-sol`, with reasoning, repository inspection, code editing and test execution. The context-window size is not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass (focused database and native suites; full local suite limitation described above) - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green (verified on immutable latest head) - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups (verified on the latest head) - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
4336b9ee02 |
fix(db): reserve transactions before checking pipeline capacity (#17831)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Interactive requests share a PostgreSQL connection pool with background work. > - Prepared queries can queue behind a slow query on a busy connection. > - Disabling pipelining lets queued work use the next free connection. > - The driver skips its transaction reservation callback when the pipeline limit is zero. > - This pull request runs that callback before the capacity check. > - Operators can test a pool without pipelining while transactions retain their connection ownership. ## Linked Issues or Issue Description **What happened?** With `PGMAX_PIPELINE=0`, postgres.js 3.4.9 skips `onexecute` in its short-circuit capacity check. A transaction then lacks its reserved connection. Real PostgreSQL reproduces `UNSAFE_TRANSACTION` through the app client and an undefined connection error through the raw driver. This prevents safe use of the existing setting to reduce query queue delays. **Expected behavior** A transaction must reserve its connection regardless of pipeline capacity. Commit, rollback, savepoints, and disconnect recovery must preserve their existing contracts. **Steps to reproduce** 1. Use the pre-change driver patch from `e61ee4b005f8e13f64af15f69cb62305dd869f82`. 2. Start disposable PostgreSQL and run the `no-pipeline` scenario in `packages/db/src/__fixtures__/postgres-connection-recovery.mjs` with `TEST_DATABASE_URL`, `DRIVER_FORMAT=esm`, and `RECOVERY_CASE=no-pipeline`. 3. Observe the transaction failure. Apply this patch and repeat with both `esm` and `cjs`. Refs #13643, which introduced the existing connection recovery patch. Searched related open and closed driver PRs. This change extends that patch and preserves its disconnect safeguards. ## What Changed - Move the transaction reservation callback before the pipeline capacity check in both driver distributions. - Add real PostgreSQL regressions for commits, concurrent transaction reads, isolation, rollback, savepoints, reserved connections, and progress on a freed pool connection. - Document the existing `PGMAX_PIPELINE=0` setting and its driver prerequisite. Keep the default pipeline limit unchanged. ## Verification - `pnpm -r typecheck`: passed. - `pnpm build`: passed. - Real PostgreSQL 18.6 in an owned, disposable Docker container: the old patch fails the new transaction scenario in both ESM and CommonJS. The corrected patch passes all 18 scenarios, including all 12 existing disconnect recovery cases. The container was removed afterward. - Local service benchmark: actual `createDb` with prepared statements, two warmed connections, two controlled advisory lock waits, and the complete three-query `getDefaultCompanyGoal` path. Two warmups and five alternating measured samples per setting. Pipeline 100: median 413.44 ms, range 413.33–416.10 ms. Pipeline 0 with the patch: median 12.22 ms, range 11.86–12.57 ms. Complete SQL, bindings, and results match. Five of five reads finish while the unrelated slow query remains blocked. These are synthetic local queue measurements, not hosted latency claims. - App client compatibility on real PostgreSQL: commit, concurrent transaction reads, rollback, nested savepoint rollback, dedicated connections, and liveness connections passed. - `pnpm test:run`: attempted, then stopped after repeated local embedded PostgreSQL startup failures caused by the host semaphore limit. This is an incomplete local full-suite run, not a pass. The focused driver checks above used disposable Docker PostgreSQL instead. The initial sandbox attempt also stopped before test execution because tsx could not bind its IPC socket. Final CI results will be recorded here. - Current head `72039f71f63530300bf241475d42e857297dfb90`: Greptile and Paperclip review are both 5/5. All three earlier findings are resolved. [CI run 40431567516](https://github.com/paperclipai/paperclip/actions/runs/40431567516) ran the complete validation matrix. The database lane passed all 265 tests, including all 18 real PostgreSQL driver regressions. One cleanup timer case timed out in server shard 6/12; its isolated real-PostgreSQL reproduction passed once and failed once, while five runs with the pre-change driver passed. That evidence is retained and does not establish the timer failure's cause. A single CI retry of the failed shard passed. The final aggregate `ci / verify` gate is waiting for a runner. Its first queued request did not start for more than 14 minutes. A guarded queue reset confirmed that it was the only unfinished job and that every test job had passed; attempt 3 restarts only that gate. No test result was replaced or bypassed. The required gate must pass before merge. ## Risks - This changes a shared driver callback order. Both module formats and existing disconnect recovery are covered by real database tests. - This PR does not change pool size, enable the setting by default, retry SQL, or change schemas. Disabling pipelining may reduce batch throughput. Operators must compare the workload before retaining that setting. - The local queue benchmark proves the mechanism. Hosted request improvement still needs acceptance after the patched image is deployed. ## Model Used OpenAI GPT-6 in Codex, with code execution and browser developer tools. The exact serving model identifier and context window are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Nicky Leach
|
e61ee4b005 | fix(runner): preserve Claude user settings trust in HOME (#17826) | ||
|
Nicky Leach
|
baec3abd67 | Authorize background project Git with explicit Connector consent (#17798) | ||
|
|
6874fed05c |
feat(adapters): refresh model catalogs and sandbox CLI pins (Oct 23 audit) (#17809)
<!-- entire-trail-link-start --> https://entire.com/gh/paperclipai/-/trails/26/changes/paperclip/26 <!-- entire-trail-link-end --> ## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Agents run on coding harnesses. Each adapter has a model catalog, and sandboxes contain pinned CLIs. > - Upstream providers released new CLI versions and new models. OpenAI retired some Codex models for ChatGPT sign-in. > - A stale catalog hides valid models. A misleading error makes operators upgrade a CLI that is already current. > - This pull request adds verified OpenCode model IDs, names Codex ChatGPT retirements in the environment Test, and refreshes the sandbox CLI pins. > - The benefit is that operators can select current models and get the correct reason when a retired model fails. ## Linked Issues or Issue Description Refs #16708 (adds Claude Haiku 5.5 to `claude_local`; this PR does not duplicate it). Refs #17245 (moves the Codex runtime to 0.163.0; this PR does not change Codex pins). **What happened?** OpenAI retired `gpt-5.5` (October 14, 2026) and `gpt-5.4` / `gpt-5.4-mini` (August 31, 2026) from Codex with ChatGPT sign-in. A saved agent with one of these models and ChatGPT sign-in now fails. The legacy environment Test says that the plan, the rollout, or an old CLI can be the cause. The Paperclip Runner Test gives generic account and runtime advice. The real cause is the retirement. **Expected behavior** Both Tests give the retirement date and the documented replacement. API-key users keep the models. **Proposed change** Record the verified retirements. Add the new verified OpenCode IDs. Refresh the sandbox CLI pins with verified digests. Record the audit. ## What Changed - Codex: add a table of models retired from Codex with ChatGPT sign-in (`gpt-5.5` → `gpt-6.1-sol`, or `gpt-6-luna` on Free and Go; `gpt-5.4` → `gpt-6-sol`; `gpt-5.4-mini` → `gpt-6-luna`). The shared `codexChatGptRetirementHint` builds the hint. The legacy CLI Test (`codexHelloProbeModelRejectionCheck`) and the Paperclip Runner Test (`testNativeRunnerAuthentication`) both use it when the backend rejects one of these models. For other failures, the Paperclip Runner Test hint stays the same. The picker keeps the models for API-key users. - OpenCode: add `anthropic/claude-haiku-5-5` and `google/gemini-4-argon` to the remote fallback catalog. The Paperclip Runner OpenCode picker also uses this catalog. - Evaluation image sandbox layer (`docker/daytona-runner`): | CLI | Before | After | | --- | --- | --- | | Grok CLI | 1.0.46 | 1.0.70 | | Gemini CLI | 0.62.0 | 0.65.0 | | Kimi Code | 2.1.1 | 2.3.3 | | Cursor CLI | 2026.10.01-e373342 | 2026.10.22-27dcf5f | | GitHub CLI | 2.102.0 | 2.103.3 | - Add `doc/adapter-model-audit-2026-10-23.md` with sources, compatibility findings, and deferred items. No model in this PR needs a newer Codex, Claude, or OpenCode runtime than the current pins. Claude Haiku 5.5 needs Claude Code 2.1.293 or later; #16708 records that floor, and this PR does not add it. ## Verification - `vitest run src/index.test.ts src/server/test.remote.test.ts src/server/acp.cli-version.test.ts` in `packages/adapters/codex-local`: 41 passed. - `vitest run src/services/native-runtime/setup-readiness.test.ts` in `server`: 98 passed. These include the two new Paperclip Runner Test cases: a retired model shows the retirement hint, and a model that is not retired keeps the generic hint. The retirement case fails on the earlier code. Run with the host's empty `ANTHROPIC_BASE_URL` unset; see the baseline note in the audit doc. - `vitest run` in `packages/adapters/opencode-local`: 87 passed. - `vitest run src/__tests__/adapter-models.test.ts` in `server`: 27 passed. - `tsc --noEmit` in `packages/adapters/codex-local` and `packages/adapters/opencode-local`: passed. - Cursor archive SHA-256 `85296944e4600a83c4b764f578d1502bf9c416db200d33a593e8f3ecabfdc294` is computed from the versioned Linux x64 tarball. The extracted binary reports `2026.10.22-27dcf5f`. On October 26, 2026, `https://cursor.com/install` still resolves this version. - GitHub CLI SHA-256 `4c806978ffce1cec70853302a56d70ac6da4f19bae7adca4e1a8ab00633dd91d` matches `gh_2.103.3_checksums.txt`. - Grok CLI 1.0.70 and Kimi Code 2.3.3 installed from npm report those versions. The flags that Paperclip uses (`-m`, `--effort`, `grok models`; `-m`, `-p`, `--output-format`, `acp`) are present in `--help`. - `git merge-tree` with #16708 (head `3a06740e9`) and #17245 (head `41f9c611f`): this PR shares no files with either PR. The two PRs can merge in any order. - Not run in the authoring environment: the full `pnpm -r typecheck` and `pnpm build` (no Rust toolchain), the full server `tsc` (memory limit), and the image build. CI runs these. ## Risks - Low risk for catalogs: the IDs are additive suggestions. No saved selection, default, or spending setting changes. - The Paperclip Runner Test changes only the hint text, and only for a ChatGPT rejection of a retired model. Check codes and pass/fail results do not change. - Sandbox CLI pins: - Kimi Code 2.2.0 and 2.3.0 remove the `kaos` MCP executor and require qualified plugin skill names. Paperclip uses neither. Kimi Code 2.3.2 stops automatic swarms in `kimi -p` and ACP sessions. - Gemini CLI 0.65.0 makes workspace settings read-only in untrusted folders. Paperclip writes home settings and sets `GEMINI_CLI_TRUST_WORKSPACE` for remote runs. - xAI publishes no Grok CLI changelog. Grok 1.0.70 behavior changes are verified only against the CLI reference and `--help`. - The pins apply only after a new image is built and promoted. The fleet sandbox image definition must get the same pins. - Rollback: revert this PR. ## Model Used - Claude Opus 5.5 (`claude-opus-5-5`), Anthropic, with tool use (shell, web research, repository search). ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge 🤖 Generated with [Claude Code](https://claude.com/claude-code) --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Devin Foleyanddevinfoley
|
cfed739f81 |
fix(external-objects): stop reporting GitHub rate limits as unreachable (#17792)
<!-- entire-trail-link-start -->
https://entire.com/gh/paperclipai/-/trails/11/changes/paperclip/11
<!-- entire-trail-link-end -->
## Thinking Path
> - Paperclip is the open source app that people use to manage AI agents
for work.
> - Paperclip shows the status of external objects, such as GitHub pull
requests and issues.
> - A background sweep refreshes these objects and resolves a GitHub
credential.
> - The sweep read a credential only from a company secret with a
well-known name.
> - An instance without that secret made anonymous GitHub API calls.
Anonymous calls are limited to 60 per hour.
> - When the limit was reached, the sweep showed "unreachable" for
objects that it could not refresh.
> - A server environment token is instance-wide, so it must not expose a
private repository that a company cannot see.
> - This pull request reads a server token for public repositories,
reports a rate limit as its own state, and shows a warning on the pill.
> - The benefit is an accurate status, fewer misleading failures, and no
cross-company private data.
## Linked Issues or Issue Description
No public GitHub issue exists. The sections below describe the bug.
**What happened?**
The board showed "unreachable" for many GitHub links, even after a
related fix was deployed.
**Expected behavior**
The board should show "unreachable" only for real network failures. It
should show a rate limit as its own state and retry later.
**Steps to reproduce**
1. Use an instance with no GitHub token company secret and no server
token.
2. Let the external-object sweep make more than 60 anonymous GitHub
requests in one hour.
3. Open a task with GitHub links.
**Paperclip version or commit**
master (commit
|
||
|
DottaandPaperclip
|
27a87c0f22 |
perf(runner): keep routine journal flushes off the server thread (#17813)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Native runners persist incoming events in the server process. > - Routine progress and usage events can cause many full journal writes. > - Synchronous file flushes stop that process from serving unrelated requests. > - Output deltas already use asynchronous writes, but other frequent events still block. > - This pull request stages routine events privately and publishes them after persistence. > - Critical lifecycle events keep their existing ordering. ## Linked Issues or Issue Description Refs: #17745, #17514. **What happened?** Page navigation, task saves, and archive requests stalled during native runner output. A post-fix CPU profile still spent about 7 seconds of a 19-second window in synchronous event journal flushes. Progress and usage events caused much of this traffic. **Expected behavior** Routine event file writes must let unrelated HTTP requests run. The durable cursor must advance only after persistence. Completion and semantic tool admission must retain their ordering. **Steps to reproduce** Run a native task that emits frequent tool progress and usage events. Send HTTP requests while its volume flushes are slow. The controlled tests hold a real file flush and send a real HTTP request during that hold. **Paperclip version or commit** Reproduced after #17745 on master. ## What Changed - Stage priority 1 and 2 event snapshots with asynchronous file writes and flushes. - Keep candidate events out of the public in-memory journal until the durable commit finishes. - Include a pending event in a command's existing durable write. Command bursts do not cause repeated event writes or business callbacks. - Reject stale candidates after an authenticated connection replaces their owner. - Keep priority 0 events, all run and turn lifecycle events, and semantic tool input persistence synchronous, including terminals labeled priority 1. - Extend slow-volume tests for progress, usage, critical events, concurrent commands, reconnects, write failures, drain, and replay. - Update the durable event delivery documentation. ## Verification - Controlled regression: the previous implementation failed the new progress and usage slow-flush cases and exposed pending delta state. The updated implementation passes all sixteen controlled cases. - All 137 durable control-plane tests and three saved-continuation cases pass. The continuation cases use the actual Rust bridge. - Review regressions failed before the correction: a priority-1 terminal took the asynchronous path and command writes did not absorb the pending event. The correction passes these tests, a two-command burst with one staged write, and a directory-flush failure after command/event rename. - An authority-boundary regression also failed before its correction. Explicit snapshot transactions now retire pending coalescing, so a later command cannot publish an older event; the test verifies replay after restart. - An initial version made lifecycle events asynchronous too. The saved-continuation tests caught missing final run receipts at completion. Keeping priority 0 events synchronous fixes those failures. This version does not claim asynchronous lifecycle support. - Repository typecheck and build pass locally after the review fixes. At the previous head `886f2b961f035071bd947b5116b1f7a2727ec17f`, run `40404783081` passed the full CI suite, including 4,024 package runner tests and 96 native-runtime tests. Final-head CI run `40409537748`, attempt 2, passed. All 46 substantive jobs and both aggregate gates succeeded. Runner lanes passed 4,032 package tests and 96 native-runtime tests. The retry carried forward the successful substantive jobs from attempt 1 and completed both aggregate gates. - Greptile reviewed final head `735245931450810246e3008231a29fd56d615f67` at 5/5 with no actionable findings. The separate Paperclip review execution ended neutrally without a new assessment. Its earlier command-churn finding is fixed and its outdated thread is resolved; the documentation and priority-1 coverage comments are also addressed. - The local wider runner suite passed 449 of 450 tests. The retained-terminal replay case fails on both this candidate and unchanged master at the same maintenance-proof check. Both local runs leave an old terminal receipt pending behind the event backlog. These failed runs remain recorded; they are not counted as passes. - The broad local test command also failed one unchanged workspace-runtime port-allocation case. That case passed in isolation. Its cause is unresolved. The broad local run was stopped after the complete CI test groups passed; no full local-suite pass is claimed. - Live staging acceptance remains pending. - The tests hold a real `FileHandle.sync`, verify an actual HTTP response during the hold, and check disk and memory cursors. They also verify command preservation and idempotent replay. ## Risks - Asynchronous writes add time for concurrent commands or reconnects to arrive. Commands absorb the event in their existing write. Revision checks prevent stale publication. Other authority changes leave the unacknowledged event for replay instead of retrying full snapshot writes. - ACKs still wait for the file flush, atomic rename, and directory flush. Write errors still fence further writes until reload. - Critical commits, JSON serialization, rename, and directory flushes still run synchronously. A command's coalesced event uses that command's existing synchronous commit. This change does not eliminate every possible source of latency. - No schema, API, dependency, or wire contract changes are required. ## Model Used OpenAI GPT-6 in Codex, with reasoning, tool use, and code execution. The session does not expose a more specific deployment model ID or context window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Nicky LeachandPaperclip
|
835966dec8 |
Prepare agent-owned environments during hiring (#17761)
<!-- entire-trail-link-start --> https://entire.com/gh/paperclipai/-/trails/1/changes/paperclip/1 <!-- entire-trail-link-end --> ## Thinking Path > - Paperclip manages agents and their execution environments. > - New agent setup saves an inactive agent before model connection tests. > - Some plugins need that agent to prepare its execution resource. > - The wizard tested without the agent ID, while the setup hold prevented preparation. > - Environment changes also discarded the selected model account. > - This change prepares the saved draft during setup and keeps normal execution held. > - Users can test the prepared resource before they finish hiring. ## Linked Issues or Issue Description **What happened?** An agent-owned plugin environment could not acquire a test lease during hiring. Changing environments also returned the user to Connect and cleared the selected account. **Expected behavior** Keep the selected account, invalidate its test, and prepare the saved agent resource in the background. Test that same agent before finishing setup. **Steps to reproduce** 1. Start a new Claude or Codex agent and select a saved account. 2. Change to an agent-owned plugin environment. 3. Try to finish setup or reconnect the account. Related: #17750 defers probes until after hire submission. This change implements preparation during the wizard instead. The related PR search and ROADMAP review found no other implementation of this setup preparation path. ## What Changed - Add the optional `requiresAgent` driver declaration. Require agent lifecycle participation when it is set. - Prepare saved drafts on creation or environment changes. Keep the setup hold and wait in `verifying` without background verification scans. - Bind tests to the saved agent. Return a warning without allocating a lease when the required agent is absent or still preparing. - Preserve the account and configuration step when the environment changes. Save the selection and invalidate the old test. - Show active preparation and support authorized retries. Keep approval, company, budget, and normal task admission controls. - Document the contract and cover lifecycle, lease, manifest, and wizard behavior. ## Verification - Server and UI TypeScript checks passed. - 112 wizard and setup helper tests passed. The shared manifest suite passed. - 123 lifecycle, policy, lease, and environment-test API cases passed after the approval-gate fix. The diagnostic and driver suites and all 80 agent adapter route tests also passed. - Token gates, module boundaries, map validation, and whitespace checks passed. - All 66 lifecycle/policy cases passed after the stopping-draft permission fix. The new real-DB route case verifies creator success, other-creator denial, and refusal to complete a hire while stopping. - Full workspace typecheck and build were attempted. Both stop at the unchanged Rust Runner step because Cargo is not installed on this host. - CI on |
||
|
Nicky LeachandPaperclip
|
c7c12dbc79 |
Support plugin panels in task sidebars (#17796)
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work.
> - Task side panels keep live work beside the conversation.
> - Plugins can already provide agent detail tabs.
> - Those tabs have no generic way to appear beside the task that is
using the agent.
> - This pull request adds an optional task-side-panel declaration to
agent detail tabs.
> - The plugin owns the label, component and availability. Core supplies
the task's running agent.
## Linked Issues or Issue Description
**What existing behavior does this improve?**
Access to plugin-contributed live views from task chat.
**Subsystem affected**
Plugin manifest validation, slot discovery, and task side-panel
UI/state.
**Current behavior**
Agent detail-tab contributions appear on agent pages. A task cannot
offer the same view through a plugin-owned availability decision.
**Proposed behavior**
An agent detail tab can declare `taskSidePanel: { availabilityPath }`.
Core substitutes the running agent into the plugin-local path. A
declared board GET route scoped to that agent returns `{ available:
boolean }`. Available contributions appear in desktop and mobile task
side panels.
**Reason and benefit**
Keep the task and its live plugin views together. Preserve plugin
ownership of provider protocols and product behavior. Core contains no
concrete plugin identifiers, routes, labels or readiness rules.
**Breaking changes**
None. The field is optional. Existing contributions remain unchanged.
Missing, disabled or unavailable plugins add no panel.
**Additional context**
Related: #17111 established the agent detail-tab and scoped API
extension surfaces. Roadmap and related PR searches found no duplicate
of this extension.
## What Changed
- Extend the manifest with optional task-panel metadata. Validate safe
plugin-local paths and require a declared board GET route with
agent-scoped authorization.
- Discover and load only opted-in agent contributions. Poll their
generic availability responses and mount the existing components with
agent context.
- Add generic panel tabs, launcher entries, dismissal and inactive-view
removal in desktop/mobile task chat.
- Reuse the existing live-run polling guard so idle tasks with a healthy
live connection stop polling.
- Add provider-neutral fixtures, schema/authorization tests, task tests
and SDK/feature documentation.
## Verification
- PASS: 301 focused tests including the full task-detail file, generic
panel discovery, persistence, slot loading and manifest validation.
Regressions cover POST-before-GET route order, mobile drawer dismissal,
recovery without focus and the absence of the hidden desktop task panel
on mobile.
- PASS: UI and shared-package typechecks; production UI build; token
gates, feature-map and module-boundary checks; diff and secrets/PII
review.
- PASS on final head: all 795 UI files / 9,186 tests, and all 95
shared-package files / 1,038 tests. Commands: `node
node_modules/vitest/vitest.mjs run --project @paperclipai/ui
--maxWorkers 3` and the corresponding `--project @paperclipai/shared`
command with local socket access.
- On the final head, server shards 7/12 and 10/12 and workspace shard
a-1/2 were interrupted by runner shutdown/communication loss. GitHub
reported infrastructure failures; no assertion failure was surfaced in
the available diagnostics. The queued summary was cancelled, then `gh
run rerun --failed` successfully requested retries for only those three
jobs. They are currently queued for CI runners; the 50 successful checks
remain passing. Interrupted attempts are not counted as passing.
[Final-head CI and
retry](https://github.com/paperclipai/paperclip/actions/runs/40396870920).
- Earlier verification exposed an empty-panel rerender loop, which was
fixed. The sandboxed combined run also failed five shared port-registry
tests because it could not bind loopback listeners; the full shared
rerun with socket access passed. These earlier failures are retained
here rather than reported as passing runs.
- Full workspace typecheck/build were attempted with pinned pnpm 9 after
resolving the initial local package-manager setup. Both stop at the
unchanged Runner Rust step because `cargo` is not installed. Full local
workspace checks are not reported as passing; final CI must provide
those broader checks.
- No live provider or plugin-specific transport acceptance is claimed by
Core's synthetic fixtures.
## Risks
- Availability is polled every three seconds only while an authorized
running agent exists. Plugins own the side effects and meaning of that
read.
- A view uses the running agent's context. It is not a separate per-task
resource.
- Optional metadata adds no backend endpoint or permission bypass. The
host requires existing board and agent authorization.
- Reverting the optional contract and UI integration removes task panels
without a data migration.
## Model Used
OpenAI Codex (GPT-6), with reasoning, repository tools and code
execution. The exact backend snapshot ID and context window size are not
exposed by this session.
## Checklist
- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have described the issue in-PR following the relevant issue
template
- [x] I have not referenced internal/instance-local Paperclip issues or
links
- [x] My branch name describes the change and contains no internal
ticket id
- [x] I have run focused tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge
---------
Co-authored-by: Paperclip <noreply@paperclip.ing>
|
||
|
DottaandPaperclip
|
718c73dafe |
fix(native): require verified remote process exit receipts (#17799)
Require fresh ESRCH or an exact matching zombie generation before treating a remote Runner as stopped. Preserve uncertain ownership through adoption and transport close, retire local controller resources after failed close, and check Node before remote launch. Bind the shared transport change in Copilot profile 62 while retaining historical profile decoding. Add red-green process, permission, launch, adoption and cleanup regressions. Thanks to @hsm-lv for the original investigation in #17073. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
Nicky LeachandPaperclip
|
4ff493fb40 |
feat(plugins): include managed skills for selected environments (#17774)
## Thinking Path > - Paperclip manages agents and their execution environments. > - Plugins can vendor company skills and provide environments. > - Environment tools need their instructions before the provider starts. > - Persistent computers also need those skills before any run begins. > - This change adds environment requirements to normal skill selection and lifecycle synchronization. > - Plugin names prevent ambiguity, and required skills take delivery precedence without deleting company data. ## Linked Issues or Issue Description **Problem or motivation** An environment can provide desktop tools while its agent receives no instructions for those tools. A first-run-only import also leaves idle computers without required skills. Same-name company skills can make provider discovery ambiguous. **Proposed solution** Let plugin environment drivers declare their managed `skillKeys`. Install missing requirements during lifecycle reconciliation or new-run preparation. Deliver updates through the durable runtime event journal. Give managed imports stable plugin names and report shadowed assignments. **Alternatives considered** ROM links do not make a skill a normal Paperclip selection. Polling duplicates the existing lifecycle delivery queue. Replacing a same-name company record would destroy independent data and assignments. **Roadmap alignment** This extends the shipped Plugin system and Skills Manager. The existing related-PR search found no duplicate contract. ## What Changed - Validate `environmentDrivers[].skillKeys` as unique references to the same plugin's managed skills. Require a ready, enabled plugin with `skills.managed` authority. - Resolve company-scoped skills before version capture. Preserve explicit pins, company edits, and saved recovery selections. Serialize concurrent first installs. - Allow lifecycle callers to request missing-skill installation through `agents.getRuntime`. Ordinary task and recovery reads do not install new declarations. Pass task-selected environment IDs through the SDK. - Add SQL migration `0378_environment_skill_runtime_hooks`. Required skill edits, effective default changes, environment changes, and plugin requirement changes enqueue durable runtime events. Reuse lifecycle delivery and retries; add no polling. - Import managed defaults as `plugin-<short-plugin-id>-<skill-slug>`, including the SKILL.md name. Preserve canonical keys. Keep same-name company records separate without counter suffixes. - Required environment skills win delivery over same-name company assignments. Preserve saved assignments and library content. Record shadowed and winning keys in run warnings, run snapshots, and plugin runtime state. Removing the environment requirement restores the ordinary assignment. - Document the contract for `driver: "plugin"`. Other drivers and plugins without requirements keep existing behavior. Existing managed copies retain edits and names until reset. ## Verification - Latest targeted local checks: environment skills (5), managed skills (9), heartbeat execution configuration (27), and SDK worker RPC host (23): **65 tests passed** (including the targeted schema-drift check). Disposable PostgreSQL exercised the migration, first install before any run, concurrent lifecycle reads, implicit edit events, inherited defaults, rollback, plugin requirement changes, shadowing, preserved records, version capture, and recovery. - Review follow-up also verifies case-insensitive/runtime-name aliases, recovery of a task-mentioned shadowed skill, preserved quoted/multiline YAML, and zero runtime events for repeated local-folder settings writes. - Migration numbering and safety checks passed. Whitespace and secret/PII scans passed. - Earlier targeted manifest validation passed (43 tests). No validator behavior changed in this follow-up. - Broad tests, typecheck, build and end-to-end verification passed in [GitHub CI](https://github.com/paperclipai/paperclip/actions/runs/40387647219) for current head `ce57f2a4d`. Greptile is 5/5 with no unresolved threads. The [commitperclip review check](https://github.com/paperclipai/paperclip/actions/runs/40387646208) also passed after restoring its required duplicate-search checkbox. - Earlier attempts found a SQL alias collision, an unparenthesized CASE expression, and fixture assumptions about bundled catalog rows and YAML quoting. These were corrected before the passing checks. The first broad CI run also exposed a schema-drift check that assumed every journal entry had a schema snapshot; it now reads the latest actual snapshot, so SQL-only migrations need no invented snapshot. A test callback replacement also affected an unrelated fixture; it was restored before the final passing execution-configuration suite. Sandbox runs could not start PostgreSQL; the reported database tests used local socket access. ## Risks - Apply migration 0378 before using the updated lifecycle contract. It adds trigger functions and event producers; it does not rewrite agent assignments or skill content. - Required environment skills take precedence over same-name assignments, including task-selected skills. Both library records remain. Run warnings identify omissions. - Company edits are preserved. Plugin upgrades do not reset installed content or rename existing copies. Explicit reset adopts current defaults. - Namespaced library slugs can be shared by distinct canonical records. Callers must use IDs or canonical keys when a slug is ambiguous. - No live provider request or deployment was performed. The downstream plugin must be built against this Core revision. Persistent-home synchronization retains its existing edit-conflict and retry behavior. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, repository tools, and code execution. The exact model identifier and context-window size are not exposed in this task context. ## Checklist - [x] Thinking path and issue description included - [x] Model and available capabilities recorded - [x] Roadmap and related PRs checked - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] Public references only; no internal issue IDs or instance URLs - [x] Targeted local tests pass; broad verification delegated to GitHub CI - [x] New behavior covered and documentation updated - [x] Current-head CI green and Greptile 5/5 with no unresolved comments --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
507dc70be2 |
fix(evals): bind hosted controller admission to the first attempt (#17523)
Bind hosted controller admission to the repository, GitHub run, first attempt, workflow revision and selected execution before credential reads or provider launch. Keep offline evidence replay independent of the current worker environment. Preserve the single-project boundary editor and refresh project-access data with agent-detail updates.
Verified head:
|
||
|
DottaandPaperclip
|
9ae4c25e7e |
perf(runner): reuse compatible protected provider packs (#17749)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Native Runner starts provider runtimes in remote execution environments. > - Sandbox images already contain qualified provider packs. > - Startup required the image pack to match the entire current app release manifest. > - An unrelated app commit could therefore cause pack verification, transfer, and installation again. > - This pull request selects protected packs by their provider interface and platform. > - The benefit is no extra pack probe, directory scan, version command, link, or upload on the compatible image path. ## Linked Issues or Issue Description **What happened?** Starting a remote task can reject a compatible image pack because its source revision or unrelated provider differs from the controller. Pack replacement added about 16 seconds in the investigated startup. Startup also repeated full content checks. **Expected behavior** Use an already qualified, protected pack when its selected provider interface and target platform are compatible. Keep the selected artifact identity for process launch and recovery. **Steps to reproduce** 1. Install a qualified provider pack in a sandbox image. 2. Start a controller built from another app commit. 3. Start a remote OpenCode task. The old complete-manifest comparison rejects the image pack even when the provider interface still works. Related repairs: #16493 reduced transfer cost. #17512 retained exact artifact authority during reattachment. This change preserves that recovery rule and changes fresh-launch pack selection. ## What Changed - Add explicit OpenCode and ACPX interface revisions to produced pack manifests. App Git SHAs and dependency pins remain provenance. - Read bounded protected-installation metadata in the existing final runner metadata query. Accept only root-owned paths that the execution user cannot write, including ancestors and manifest. Reject writable aliases. - Select and prepare the pack in server-owned environment setup before constructing the existing transport. Bind provider launch paths and digests from the selected installation. Save that identity for exact reattachment. - Keep full content verification for uploads, writable installations, and legacy packs. Reuse successful verification within one startup. Keep older attached runners compatible with the optional metadata request. - Seal image packs and document the one-time image refresh that enables reuse across later compatible releases. Keep selected ACPX agent qualification checks. ## Verification - Red/green regression: the old code rejects a protected pack from another app revision with `manifest mismatch`; the fixed path selects it without executing verification commands. - Native startup regression: select the installed pack's own Node digest with one existing runner metadata request and no pack verification or version commands. - Targeted pack, artifact, recovery, execution-driver, and Rust metadata tests pass. Incompatible interfaces/platforms, malformed identities, writable roots, path aliases, and exact reattachment are covered. - Final revision `91dc1676ff67bdbe5866c8586daac10240530cca`: workspace typecheck and build pass. All 53 current-head checks are successful, with no pending or failed checks. Greptile is 5/5 with no open findings. - The broad local `pnpm test:run` attempt reported a workspace-runtime conflict test failure; its isolated rerun passed. The broad run was stopped after that failure. Full CI now passes, including the server test shards. This is not a claim that the full local run passed. - Linux filesystem evidence: the production metadata functions read a real root-owned preinstalled pack as a non-root user in 16 microseconds median and 20 microseconds p95 across 100 warm reads. This measures metadata admission, not model or hosted response latency. - All 910 affected server tests and 57 Copilot attestation tests pass locally on the final implementation. Full CI covers the unchanged Runner transport. No provider qualification hash was refreshed. - Built the actual Linux daemon at `416e7ec7425d0d3565b4a5e1808002d650a60414` and ran its metadata command as UID 65534. The protected installation returned its real manifest with `immutable: true`; writable alias and missing-pack cases returned `immutable: false`. This is metadata/permission evidence, not provider execution. - Hosted rollout and a hosted end-to-end latency measurement have not run. ## Risks - Existing images need one rebuild to publish the compatibility declaration. Packs without it keep the legacy content-verification path. - Fast admission relies on image publication producing a qualified protected tree. Root execution and task-writable installations retain full verification. - Bump an interface revision when required runner/pack behavior changes incompatibly. Deploy a new image to roll out compatible runtime fixes or upgrades. - This removes provider-pack preparation delay. Other runner startup, context, model, and finalization costs remain separate. ## Model Used OpenAI Codex, configured model `gpt-6.1-sol`, reasoning effort `xhigh`, with tool use and code execution. The context-window size is not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
f1922d95e1 |
test(runner): qualify the OpenCode discovery fixture on Linux (#17757)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Runner tests must exercise the same verified launch boundary as real providers. > - The assigned-tool test from #17740 used an executable JavaScript script as its OpenCode fixture. > - Linux launches verified executable bytes from a sealed, unlinked memory file. > - Node cannot reopen that deleted script path. Cleanup then hid the startup error. > - This pull request uses the native fixture wrapper already used by the adjacent OpenCode integration test. ## Linked Issues or Issue Description Refs #17740. Related fixture qualification work: #14529. **What happened?** The merged-source Cloud readiness job failed twice in the new assigned-tool integration test. Linux could not start the script fixture from the verified executable file. The later cleanup error hid the original `ENOENT` error for `/memfd:paperclip-verified-launch (deleted)`. **Expected behavior** The test must pass through the real Rust Runner, OpenCode proxy, and HTTP MCP bridge on Linux. It must also report the original error if cleanup fails. **Steps to reproduce** Run `vitest run src/live/runnerd-opencode-assigned-tools.test.ts` in the Runner package with its Linux binary and built proxy. The original fixture fails during startup. The native wrapper passes the assigned-tool assertions and durable close. ## What Changed - Launch the existing JavaScript server fixture through a native wrapper. - Qualify an owned Node copy on Linux with strict file permissions. - Bundle the current production proxy into the test's temporary directory. The test no longer skips when shared `dist` output is absent. - Preserve both the original error and a cleanup error. - Remove the temporary fixture even when setup fails. ## Verification - Reproduced the failure in a local Linux container with the exact merged Runner binary from the published Core image. - The same Linux test passes after this change. It checks the assigned discovery call, the completion call, and successful durable close. - The updated test passes on macOS. - A real Linux compiler-failure probe leaves zero fixture directories behind. - `pnpm -r typecheck` and `pnpm build`: passed. - Runner TypeScript checks pass after the cleanup review change. - Full technical CI passed on `6ee1865af3310f13523b11407f9e41879832f543`. - Fresh Greptile review: 5/5, with no unresolved findings. - Fresh Paperclip Review published 5/5 on `6ee1865af3310f13523b11407f9e41879832f543`, with no actionable findings. - The optional Entire assessment remains pending because this branch has no assessed Entire Change. The maintainer approved merging after CI and both reviews pass. - The full local test suite was not repeated. Earlier attempts hit host PostgreSQL semaphore limits. CI provides the full suite check. ## Risks Low risk. This changes one test fixture. Runtime behavior and launch verification stay the same. The test now needs the C compiler already required by the adjacent native fixture tests. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, tool use, and code execution. The runtime does not expose the exact backend model ID or context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
8c702d59db |
feat(evals): package a portable Runner controller runtime (#17506)
Package a verified portable Runner eval controller with complete bounded file receipts, copied dependency closure, private evidence replay and symlink-safe CLI entrypoints. Preserve fail-closed cost holds, native trace receipts and all parent security repairs.
Verified head:
|
||
|
DottaandPaperclip
|
5061ab7b35 |
perf(ui): speed up task and inbox cold opens (#17756)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Task lists and inboxes are the main entry points to that work. > - Cold opens fetch overlapping task lists, task descriptions, and run details before useful rows settle. > - Independent responses can insert rows above content that a person has started to read. > - This pull request loads a small task index first and adds details only for visible rows. > - It keeps row order and geometry stable while those details arrive. ## Linked Issues or Issue Description **What existing behavior does this improve?** Cold opens of Tasks and Inbox, including the shared navigation badge. **Subsystem affected** Server list reads, shared response contracts, and the React UI. **Current behavior** The badge fetches full task bodies. Mine fetches several overlapping lists. Historical diagnostics and row metadata add work before the first useful list. A fast failed-run response can appear before task rows, then move down when tasks arrive. **Proposed behavior** Load the selected inbox source and its required context together. Use an index without task descriptions or recovery diagnostics. Fetch diagnostic details for up to 50 visible rows. Keep skeletons, recovery indicators, external status, and live metadata in fixed space. **Reason and benefit** Reduce work on a cold open and prevent loading-induced row movement. On the same 100 local fixture tasks, the compact response was 262,441 decoded bytes and the index was 110,909 bytes. This comparison excludes compression and is not a page latency benchmark. **Breaking changes** Existing full and compact responses remain available. New index/row projections are opt-in and marked as partial data. Mine keeps its existing most-recent-100 display limit, now selected on the server. Existing filters and group preferences remain in the client. No database migration. **Additional context** Refs: #17593. Related: #13095, #17428, #13803. This change addresses collection overfetch and layout stability. It does not replace the blocked-inbox query work or the database activity-index work. ## What Changed - Add authorized index and row projections, with bounded task and parent ID batches. - Replace badge list downloads with unread IDs and aggregate counts. Preserve actor privacy, dismissals, and local archive behavior. - Load only the selected inbox source, plus bounded parent and failed-run context. Return only relevant failures from the bounded run window. - Batch hierarchy reads, including at most one authorized child per collapsed parent in one request. Collapsed markers do not consume the expanded row budget, and partial descendant counts are not shown as totals. Enrich visible rows without replacing membership, ordering, titles, or read state. - Reserve row geometry on desktop and mobile. Reveal the initial inbox only after its sources settle. Preserve legacy task-list reads. - Start collection module downloads during access checks. Keep partial rows out of the task-detail cache. - Keep ETags and concurrent request sharing for the new projections, but skip the short server response TTL so archive undo sees fresh state. ## Verification - On final head `ef6afb2a56`, the full CI workflow passed, including build, typecheck, server/workspace/runner suites, and all eight browser shards. A server shard reported **1,489 passing tests** before runner shutdown canceled the job; its single infrastructure retry passed. Both Greptile and Paperclip Review are **5/5**, with zero unresolved review threads. - The optional external **Entire Gates** check remains `unassessed` because it has no active assessed Change for this comparison. Entire is not set up in this checkout. The repository ruleset requires `ci / verify` and `ci / e2e`; both passed. No gate configuration was changed. - Passed workspace typecheck after the final API batching change. The full production build also passed on the final commit; the exact-retry E2E run built the shipped UI. - Passed token gates and map validation. The final batch-preview pass has 91 passing route/hierarchy/component tests, including privacy, request bounds, 30 collapsed parents in one read, and a 120-child parent alongside a sibling. - Passed the 166-test focused UI/inbox/badge run, 47 additional UI regression tests, and 34 issue-list route tests. The latest regressions cover low-trust agent badge IDs, cold company selection, 30 description-only Mine matches with archive actions, retrying optional row details, and collapsed/expanded siblings. - The full local `pnpm test:run` spanned the rebase and reported failures in five suites before it was stopped; it is not claimed as a complete pass. A fresh serial rerun of all five suites passed **452/452** tests (tool gateway, native resume, project repositories, workspace recovery, and runner selection). Full current-commit coverage is checked by CI. - Real browser: delayed index and row-detail responses on Tasks and Mine. The first placeholder and first row had the same origin and 44-pixel height. Twelve rows on each page had zero position, height, or title-width change after recovery details arrived. - Real browser: reviewed long titles and recovery status at a 390 by 844 viewport. Task text stays on one line and recovery uses a compact indicator. - Three fresh document loads per page on final commit `260893bf92`, with HTTP cache retained on a warm local development server: Tasks 1.031–1.190 seconds; Mine 0.999–1.193 seconds. Browser-control overhead is included. These values must not be compared directly with staging latency or the earlier local samples taken during a full test run. - The browser archive test exposed stale results after Undo. The new projections now bypass the server TTL; immediate-mutation regression tests pass. The final browser retry passed: Archive followed by Undo restored exactly one row and the badge count. - CI exposed a hook-order crash when company selection resolved after the first inbox render. The hook now runs before the empty-company return. All seven affected exact-retry and legacy-continuation browser scenarios pass locally after the fix. - Real browser review regressions: 97 description-only source matches retained Archive actions; deliberately failed optional row reads displayed a retry notice while all 12 observed row positions and heights stayed unchanged. Retry restored the recovery indicator. Network fault injection was removed. - The final production build and workspace typecheck passed. Final focused verification passed 153 UI/helper tests and 26 heartbeat privacy route tests. Final review regressions cover newer diagnostics arriving during the index read, reloading diagnostics when the same task ID changes version, optional source failure/retry without hiding available rows, and required external-status filter failure. The latest-failures route test verifies bounded selection, agent filtering, success suppressing an older failure, and private-run redaction. - Real browser: deliberately failed the run source, then retried it. Twelve available task rows kept the same position and 44-pixel height before, during, and after retry. No loading skeleton replaced them. Fault injection was removed. - Hierarchy review decision: the cold-open barrier stays because inserting ancestors or expanded children after revealing index rows would cause layout shifts. Existing hierarchy context now survives subsequent requests, with a regression proving a new ancestor read does not blank the list. The bounded recursive tree read can still add latency for unusually deep expanded trees. - Tasks now handles a failed external-status filter with an explicit error and Retry, suppressing a false empty state. All 52 Tasks component tests pass, including failure-to-success retry. One existing large-tree test timed out while workspace typechecking ran concurrently; the whole component suite passed when rerun without that competing work. Workspace typecheck, the full production build, and token gates passed for this fix. ## Risks - Saved filters and grouping still operate on the existing bounded collection. This is not a new global server pagination model. - Fixed row slots reduce available title width. Long titles remain accessible through task navigation. - Recovery and related diagnostics arrive after core rows. Their responses cannot replace ordering fields, and stale versions are rejected. - Index/row requests now revalidate against current data instead of a short server TTL. They still share concurrent reads and use ETags. - This has not been deployed by this pull request. Staging latency must be measured after deployment under comparable conditions. ## Model Used OpenAI Codex, GPT-6 family. The exact runtime model ID and context-window size are not exposed in this session. Used code editing, terminal execution, and real-browser diagnostics. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
313c1db2c3 |
fix(evals): share bounded request policy with qualification (#17765)
## Thinking Path > - Paperclip manages agents and their work. > - Product E2E evaluations send model work through a gateway with cost limits. > - The gateway checks the request body before it reserves money or contacts a provider. > - A qualification probe needs that same check before it records a spending-cap denial. > - Namespace support already exists. Qualification must share that policy without sending a paid request. > - This pull request exposes the pure request validator and bounds the supported local tool definitions. > - The benefit is one request policy for qualification and forwarding. ## Linked Issues or Issue Description Refs: #17452, #17469. **What happened?** The request validator was private to `forwardEvalRequest`. A qualification caller could not reuse the request policy before admission. A separate check could accept a request body that forwarding rejects before a reservation. **Expected behavior** Qualification and forwarding use the same pure validator. One-level OpenAI namespaces with local function or custom tools remain supported. Invalid tool definitions fail before accounting or transport work. **Steps to reproduce** Use the synthetic native-daemon request fixture in `eval-gateway.test.ts`. It has seven local function tools and one namespace with five local function tools. Call the shared validator, then forward through the fake provider. Run the malformed namespace and tool-count cases to check rejection before side effects. The namespace behavior follows the [OpenAI function-calling guide](https://developers.openai.com/api/docs/guides/function-calling). This change preserves the one-level contract introduced in #17452 and hardened in #17469. ## What Changed - Export `boundedEvalPayload` with credential-free route fields. `forwardEvalRequest` calls that function. - Keep one namespace level with function or custom leaves. Require exact leaf types, nonblank string leaf names, and a closed namespace wrapper. - Limit OpenAI tool definitions to 1,024 per request. Count namespace wrappers and their children. - Add native-daemon-shaped forwarding coverage, malformed and hosted tool rejection, count and depth boundaries, and Anthropic compatibility checks. - Document the shared qualification policy in the Product E2E security guide. - Compile the real observability validator in the test parent. The owned CLI fixture imports JavaScript, so its process group does not include a test-only TypeScript compiler service. The parent resolves the compiler from the root package, which declares that dependency. Production cleanup checks and attempt deadlines stay intact. ## Verification - `pnpm exec vitest run --config tests/runner-e2e/vitest.config.ts eval-gateway.test.ts cost-ledger.test.ts provider-usage.test.ts operating-overhead.test.ts operations.test.ts`: 173 passed. - `pnpm test:e2e:runner:typecheck`: passed. - `pnpm -r typecheck`: passed. - `pnpm build`: passed. - Separate Node support checks: 137 passed. - `pnpm test:e2e:runner:unit`: 4,271 passed, two skipped, four failed in unchanged Daytona image and Pi catalog contracts. The same four failures reproduce on untouched base commit `a67a18c171816b097f46eae78fc58c7e55c9c565` (79 passed, four failed in those files). - `pnpm test:run`: interrupted and incomplete after required CI completed. Initial database startup failed because this checkout omitted the embedded PostgreSQL package postinstall aliases. The package hydrator restored them, and 14 real database checks passed. Earlier failures and the full output remain preserved; this local full run is not counted as passing verification. Its owned process group exited after SIGINT, with no owned survivors. - An offline full native authentication probe used the real Rust daemon and pinned Codex 0.160.0 Darwin SDK. The same captured request failed the older operations policy before reservation and reached a mock reservation with this policy. The OS sandbox allowed loopback only. It used synthetic credentials, retained no raw model payload, and made no provider calls. - Controller fixture suite: 70 passed. Three cold comparisons each confirmed that the old child started an owned compiler service and the corrected child did not. On Darwin the old compiler also exited cleanly, so this does not prove the exact historical Linux process race. Fresh Linux CI is required. - Independent review passed. The private qualification integration passed 14 exact-source compatibility checks and 67 shared-source tests without provider calls. - Required GitHub CI checks passed on the current head. The separate Entire approval gate still awaits an eligible reviewer; the delegated author account's explicitly AI-assisted approval does not satisfy it. ## Risks Requests with more than 1,024 OpenAI tool definitions or extra namespace wrapper fields now fail closed. New provider schema fields will need review. Anthropic tool handling, model prices, billing restrictions, output caps, and reservation ordering remain unchanged. This offline request-policy proof does not establish a completed model turn or Linux or Cloud onboarding acceptance. ## Model Used OpenAI Codex, GPT-6 family. The exact serving model identifier and context window are not exposed in this session. The model used reasoning, tools, and code execution. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
125ddcabe5 |
perf(runner): keep event journal writes off the HTTP thread (#17745)
## Thinking Path > - Paperclip lets people manage AI agents and their work. > - The HTTP server also owns durable Runner state. > - Each committed event rewrites and flushes that state on the HTTP thread. > - Large journals and slow writes delay unrelated navigation and saves. > - Event acknowledgements must still wait for durable storage. > - This PR stages output writes asynchronously and joins them before writer replacement. > - The board can serve requests while Runner output waits for storage. ## Linked Issues or Issue Description Refs: #17507. Related: #17514 reduces writes per event burst. This PR moves individual output writes and file flushes outside the HTTP thread. **What happened?** A live CPU profile spent about three seconds of a 13-second sample serializing and synchronously writing Runner state. Representative journals were about 9 MB. Unrelated task requests had multi-second delays. **Expected behavior** A slow output-journal write must leave unrelated HTTP requests responsive. An event must not be acknowledged before it is durable. **Steps to reproduce** Hold an authenticated output event's file flush in the durable protocol test. Send an unrelated HTTP request during that flush. Also save a command, reconnect, or retire the controller during the write. ## What Changed - Stage `item.delta` snapshots with asynchronous file writes and file flushes. Keep lifecycle, control, authentication, and tool events synchronous so transport readers observe durable boundaries. - Fence publication by saved revision. An older snapshot cannot replace newer commands or connection state. A failed write fences the writer. - Write compact JSON and share private-file checks. Existing journals remain readable. - Stop external ingress before joining admitted writes. Checkpoint only after retirement succeeds. Close reports a failed detach consistently and releases notification consumers. - Add real HTTP and Rust transport cases for held writes, reconnects, commands, failure, restart, close, and timeout. - Add immutable Copilot source identity v61 for the changed shared transport. Preserve v60, historical decoding, exact-current launch checks, and qualification flags. No new live Copilot qualification is claimed. ## Verification - Head `6f5c92e778`: workspace `pnpm -r typecheck` and `pnpm build` passed. All CI gates passed in [run 40356159484](https://github.com/paperclipai/paperclip/actions/runs/40356159484), including both full Runner shards. Fresh Greptile review is 5/5; all findings are resolved. - Current-base focused verification passed 300 cases across protocol, real Rust restart/continuation, permission, backend, and immutable-profile suites. The complete transport suite also passed 286 cases before the base refresh. - Earlier iterations exposed lifecycle visibility, late writer publication, and close-order failures. Controlled regressions reproduced each before repair. CI also found missing test-peer drain methods and a stale source fingerprint; both are corrected. An older broad local test run was stopped after its source was superseded. No full local-suite pass is claimed; full coverage passed in isolated CI. - A synthetic 9 MB journal comparison used 20 writes per sample and concurrent HTTP probes, with alternating baseline/treatment order. HTTP median changed from 31 ms to 2–3 ms; p95 changed from 51–85 ms to 12–21 ms. Parsed output was equal. The asynchronous loop served 125–153 probes versus 19–20, and took longer overall (992–1069 ms versus 675–678 ms). This proves local I/O responsiveness, not deployed application latency. - Deployment and live save/archive acceptance remain pending. ## Risks Persistence interleavings affect restart authority. The tests cover concurrent commands, replacement, failed storage, and reopening the journal. JSON serialization, atomic rename, directory flush, and control-event writes remain synchronous. This change still rewrites whole journals. Incomplete retirement retains recovery evidence and fails the close. A new exact-current Copilot source identity can reject an older launch identity; historical decoding remains available. No migration is required. ## Model Used OpenAI GPT-6 (Codex), with reasoning, tool use, SSH and browser diagnostics, code editing, and test execution. The exact serving model suffix and context-window size are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
063a498b54 |
Bound native OpenCode assigned-tool catalog projection (#17740)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Native agents use assigned connection tools through a governed gateway. > - OpenCode received the gateway catalog directly, outside the bounded server tool projection. > - A large catalog can exceed a provider grammar compiler's capacity even when it fits the transport limits. > - Synthetic requests reproduced the Fireworks grammar timeout on the configured DeepSeek model. > - This pull request gives OpenCode two assigned-tool discovery and call schemas through the existing server relay. > - The gateway keeps the same grants, validation, approvals, revocation, and audit rules. ## Linked Issues or Issue Description **What happened?** A native OpenCode run could fail during provider grammar compilation before its first assigned-tool call. OpenCode advertised the full assigned gateway catalog. A synthetic catalog of 741 tool schemas reproduced `Grammar processing took too much time` on Fireworks through OpenRouter with `deepseek/deepseek-v4.1-flash-1017`. The same model and backend accepted a compact two-tool discovery/call catalog. This reproduces the mechanism; it does not replay a historical private request. **Expected behavior** Assigned-tool discovery must keep provider grammar size bounded. Every call must retain the run's original gateway binding and permissions. **Steps to reproduce** 1. Give a native OpenCode run a large assigned connection catalog. 2. Use a provider that compiles tool schemas into a constrained output grammar. 3. Observe the full catalog in the provider tool surface before this change. 4. After this change, observe only `paperclip_search_assigned_tools` and `paperclip_call_assigned_tool` for assigned tools. Discover and call a permitted tool through its pinned alias. **Paperclip version or commit** Source checkout based on current `master`. **Deployment mode** Native local or sandbox execution with OpenCode 1.18.34. Related work: Refs #14258 (discovery paging), #16678 (local adapter MCP launch), and #17688 (gateway capability advertisement). This change covers native OpenCode catalog projection. ## What Changed - Route native OpenCode assigned tools through the existing server-owned authenticated relay. - Use bounded discovery/call schemas for OpenCode, including small catalogs. Preserve transport capacity checks. - Keep assigned gateway credentials on the controller. Ignore gateway and credential values in tool arguments. - Preserve exact declared MCP operation names through admission and driver dispatch. Resolve transport prefixes only against the closed catalog. Reject unknown or shortened names. - Refresh former OpenCode direct-catalog and shortened-name checkpoints. Keep the existing fingerprints for other providers. - Add local and remote wiring, catalog capacity, alias, scope, work-mode, approval, and revocation tests. - Allow the native test files to use a validated dedicated external test database when embedded PostgreSQL is unavailable. - Document the assigned-tool execution contract. - Bound OpenCode discovery pages to 32 KiB and large-schema chunks to 16K characters. Keep other harness budgets unchanged. - Test medium pages and escaped large schemas through the actual HTTP MCP bridge on the explicit Runner testing surface. ## Verification - Red before the catalog fix: OpenCode did not create the server relay, and forced compact delivery advertised 1 or 741 direct schemas. - Red before the alias repair: HTTP MCP changed declared operation names, and the driver sent `search_assigned_tools` instead of the authorized `paperclip_search_assigned_tools`. - Current repair tests: 112 OpenCode/proxy tests pass, plus one actual Rust/proxy/HTTP MCP regression. Two checkpoint regressions fail with the former scoped revision and pass with the new revision. Other providers keep their prior fingerprints. - The combined QA candidate with this exact PR head passes all 916 native tests and all 113 transport tests on supported Node 24.22. An earlier Node 26.4/macOS run passed the 113 assertions but had an uncaught networking error. That non-clean run remains recorded; its stack matches [the upstream Undici defect](https://github.com/nodejs/undici/issues/5544). No production workaround or error suppression was added. - `pnpm -r typecheck`, `pnpm build`, and `git diff --check` pass on `bc41ba72be2195eabeb0d7727227ee0cd02e8532`. The maps structural check passes. - The local `pnpm test:run` attempt stopped after host embedded PostgreSQL initialization failures. The relevant database suites pass with isolated PostgreSQL 18. All repository CI gates, including aggregate `ci / verify`, pass on the current head. - Synthetic live provider comparison: 741 invented schemas produced the exact Fireworks grammar timeout in 20.9 seconds. Two compact invented schemas produced a discovery call in 585 ms on the same model and backend. - The first hosted QA attempt exposed the alias failure described above. On the patched source, a real local Rust runner → proxy → OpenCode 1.18.34 → exact DeepSeek smoke completed discovery, one synthetic gateway read, and `paperclip_finish`, then emitted `turn/completed`. It used the production compact projection and unchanged Rust allowlist. - Two earlier local smoke attempts exposed missing terminal authority and completion-contract fields in the smoke harness. Those failures remain recorded. The corrected ordinary bindings produced the clean result. - Fresh Greptile: 5/5 on `bc41ba72be2195eabeb0d7727227ee0cd02e8532`, with no new actionable findings or unresolved threads. - Hosted QA publication and lifecycle acceptance remain pending. The separate deployed-agent `Paperclip Review` check is also pending; it follows a task that failed on the earlier deployed runtime. - Local OpenCode/OpenRouter acceptance at `bc41ba72`: 530 synthetic tools were authorized and installed. A simple model-ID reply completed in 17.3 seconds with no app calls. A discovery task completed in 37.6 seconds, used the search/call relay, made one upstream call, and returned the exact synthetic value. This is local source acceptance, not a staging sandbox timing comparison. - The large-result review finding has a failing-before/passing-after HTTP MCP regression. Five medium schemas now page without invalid JSON. A large schema with escaped text reconstructs exactly through bounded chunks. All 35 assigned-tool tests and the server typecheck pass after this follow-up. - Final head `d84ce4856a6e65a35ad5fb8b561492598b525f4e`: all CI gates pass. Greptile and Paperclip Review both rate the change 5/5. No review threads remain open. The final full build, server typecheck, and 35 assigned-tool tests pass locally. ## Risks - OpenCode uses discovery before it calls an assigned tool, including for small catalogs. This can add a model turn. - An existing OpenCode provider checkpoint rotates once because it retains the former tool declarations. Paperclip task history remains available. - The historical filtered catalog was not captured. The synthetic comparison proves the failure mechanism, not the exact cardinality of an earlier private request. - No model, provider route, permission, or automatic retry policy changes. ## Model Used The bounded-result follow-up also used OpenAI Codex (GPT-6 family). The exact backend model ID and context-window size were not exposed for that session. It used reasoning, code execution, and tool use. OpenAI `gpt-6.1-sol` in Codex, with reasoning, tool use, code execution, and source inspection. The runtime does not expose the context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
5823e9a664 |
fix(runner): preserve disabled Codex web search (#17744)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Native Codex agents use the saved agent settings to prepare each provider launch. > - Onboarding saves web search as off, but native configuration conversion removes that setting. > - Codex 0.160.0 includes a hosted web-search tool by default, even in an authentication probe that asks it not to use tools. > - A gateway that permits only priced function tools can reject that request before budget admission. > - This change preserves the explicit restriction through native execution and disables web search for authentication probes. ## Linked Issues or Issue Description **What happened?** A native Codex authentication probe can include hosted web search. Normal task execution also loses a saved `search: false` setting. A restricted Responses gateway can reject these requests before it makes a model call. Related repairs: Refs #17445, Refs #17529. The roadmap and related PRs were checked; no duplicate repair was found. **Expected behavior** Authentication probes check the selected account and model without hosted web search. A saved search-off setting remains effective in native task launches, prepared launches, and session recovery. Existing behavior for an absent setting remains unchanged. **Steps to reproduce** 1. Create a Codex agent with search off and choose Paperclip Runner. 2. Use a Responses gateway that rejects hosted tools. 3. Inspect the request from the native authentication probe or task launch. Before this fix, the pinned Codex SDK includes `web_search` despite the saved setting. ## What Changed - Preserve an explicit search-off setting when native agent configuration is built and converted. - Carry the restriction in the durable native input and provider launch requirements. - Apply the restriction to authentication probes and normal launch paths, including custom arguments and recovery. - Reject prepared launches that discard the required restriction. Keep gateway billing rules unchanged. - Admit only the restrictive literal value in the strict plugin task API. Exercise the real plugin setup path. - Add regression coverage and document the policy. - Add the immutable Copilot v60 source identity required by the changed shared files. Preserve v59 and all original live evidence. Copilot authentication probes keep their prior launch requirements. ## Verification - Regression cases were run against the unfixed source and failed at the configuration and launch boundaries. - An isolated capture using the real Codex 0.160.0 app-server reproduced hosted-tool rejection. Disabling web search reached a mocked budget reservation. The helper and child processes permitted only loopback network access and used synthetic credentials. It made zero upstream calls. - Focused checks passed: 284 server cases, 61 native form cases, 198 backend/security/launch/parser cases, and 36 authentication/launch cases. A separate complete transport/launch/parser run passed 393 cases. These groups overlap. - Head `617b98f` passed workspace typecheck/build and 28/28 plugin contract and actual Runner probe tests with no skips. Final rebased head `20ff0c1` passed all current-head CI checks, including full isolated test lanes, and fresh Greptile and Paperclip reviews at 5/5. - The additive Copilot identity update passed 134 compatibility/policy/protocol checks and nine authentication probe checks. The real pinned Codex SDK capture was repeated after the final policy change: all five cases passed with zero upstream calls. - Two local full-suite attempts stopped after native PostgreSQL loader failures. Seven local dependency aliases were restored, and an isolated cluster initialized successfully. A third run later encountered host PostgreSQL semaphore exhaustion and was interrupted before source-profile regeneration. Its zero exit code after SIGINT is explicitly not a full-suite pass. Final full-suite verification runs in isolated CI; the local attempts remain recorded. - The first CI run exposed a missing strict plugin API field. Its five failures were reproduced locally, then passed after repair. The next run passed every underlying job except the source-identity guard. That guard now binds an additive identity with v59 retained. Final-head CI and both fresh reviews passed; addressed review threads are resolved. - The offline capture runs on Darwin ARM64. It does not prove the contents of a prior unretained Linux request or complete Cloud onboarding acceptance. ## Risks - The restriction must survive prepared launches and resumed sessions. Regression tests cover those boundaries. - Existing absent settings retain their prior behavior. This change does not add native search-on support or loosen gateway admission. - Older durable inputs without the optional restriction must remain valid. - Copilot now uses profile v60 for exact launch admission. Historical v59 inputs remain readable. The new declaration adds no live Copilot qualification and does not change its executable or admission flags. - A deployed Cloud acceptance run is still required after this source repair. ## Model Used - OpenAI GPT-6 through Codex, with reasoning, code execution, and delegated review. The exact backend model identifier and context-window size are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
ad861a7f94 |
fix(runner): allow bounded controller admission before welcome (#17735)
## Thinking Path > - Paperclip manages AI agents and preserves their work across execution failures. > - Native Runner authentication exchanges a one-use bootstrap ticket for a connection lease. > - The controller must admit the connection and persist that lease before it sends welcome. > - Runnerd waited only two seconds for welcome, even when its connection setup budget allowed more time. > - A slow valid welcome therefore looked like a consumed-ticket failure before provider work could start. > - This pull request uses the existing ten-second connection setup budget for welcome. > - The benefit is reliable startup during bounded controller delays, with the same authentication and replay checks. ## Linked Issues or Issue Description **What happened?** A valid native Runner bootstrap can fail while the controller completes admission. The runner then reports `bootstrap connection failed closed; provide a fresh one-use ticket`. A later recovery attempt can report an indeterminate authority and hide the startup error behind a finalization deadline. **Expected behavior** A valid welcome can arrive within the connection setup budget. An earlier execution or reconnect deadline still ends the wait. A missing or invalid welcome never permits bootstrap-ticket replay. **Steps to reproduce** 1. Run the real Runner WebSocket transport against an authenticated controller. 2. Hold the welcome for three seconds after the client proof arrives. 3. Before this change, authentication fails with `WebSocket authentication message deadline elapsed` and `bootstrap_maybe_consumed: true`. 4. With this change, the runner accepts the bound lease and restores the normal read timeout. **Paperclip version or commit** Reproduced against `a1f77821b9` before the production change. **Agent adapter(s) involved** Native Paperclip Runner. The transport is shared by its provider drivers. Related: #17401 reduces repeated work in warm-recovery admission. This change covers delayed welcome delivery for fresh authentication too. It does not replace that optimization. #17594 addresses a separate lifetime failure after controller outages. ## What Changed - Use the ten-second connection setup budget for the authenticated welcome. - Keep the execution/reconnect deadline as an earlier bound. - Add real WebSocket regressions for a delayed welcome and a shorter lifecycle deadline. - Add a real Runner process test proving a delayed admission starts one provider and keeps it through two turns. - Document the startup contract and its recovery verification scope. ## Verification - Red: the delayed-welcome regression failed against unchanged production code with a message deadline and `bootstrap_maybe_consumed: true`. - Green: all 402 Runner core library tests pass. The suite includes authentication, ticket fencing after a lost welcome, identity rejection, replay, and no repeated provider effects. - Map and module checks pass. - Workspace `pnpm -r typecheck` and `pnpm build` pass. The new process integration test passes (one provider process, one thread, two turns, suspended with an empty outbox). - Live acceptance at this branch's production Runner build passes on macOS arm64 with pinned OpenCode 1.18.34 and `openrouter/deepseek/deepseek-v4.1-flash-1017`: delay admission by three seconds, create a file, suspend, resume the same provider session under a fresh run authority, read the first file, and write a second file. Both useful outputs match exactly; final runner state is suspended. - Preserved setup failures: a non-pinned executable was rejected, and the debug binary exceeded cold provider startup admission. The pinned executable and release binary passed. The first harness also tried to reuse a completed task object; the follow-up used the normal resume path. Exactly two paid model turns ran. - The full local `pnpm test:run` was attempted and stopped after the Mac exhausted PostgreSQL semaphore capacity (`could not create semaphores`); it is not a local full-suite pass. - Current-head CI is green at `df5d5494c9f9f444114aaf12e095a554110d2e14`: all checks passing or non-applicably skipped, zero failures or pending checks. The additional Superagent security scan triggered by ready-for-review also passed. These include all general/serialized test shards, Runner Rust/Vitest, E2E, typecheck, build, and the final verify gate. Greptile reviewed that exact commit at 5/5 with no findings or inline comments. - This proves the actual native Runner/provider path locally. It does not prove rollout or recovery on a hosted Linux/amd64 execution environment. ## Risks - A stalled welcome can take up to eight seconds longer to fail when the configured lifecycle allows it. Shorter execution/reconnect deadlines still take precedence. - Authentication proofs, credential expiry, identity binding, and one-use-ticket fencing remain enforced. - This does not repair provider-side grammar errors or authorize replay of an indeterminate historical run. Recovery of existing work requires the normal stopped-owner and retained-state checks. - No schema or migration change. A new Runner binary must reach the execution host before this change affects remote runs. ## Model Used OpenAI GPT-6 through Codex, with reasoning, repository inspection, code execution, and browser tools. The exact serving submodel identifier and context window were not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
bbe1016d81 |
feat(evals): retain bounded native tracing in the Runner controller (#17493)
Retain bounded native tracing in the Runner controller with attempt-bound relays, private export receipts, offline cost replay, conservative holds and bounded failure diagnostics.
Verified exact head
|
||
|
DottaandPaperclip
|
3f91d58dd5 |
fix(ci): provide Copilot auth only to protected paid eval cells (#17743)
## Thinking Path > - Paperclip manages AI agents and their work. > - Paid runner evals test real provider integrations. > - Copilot cells use the saved-token connection path. > - The two trusted workflows did not pass the Copilot credential to those cells. > - This PR supplies it only when the selected cell needs it. > - Existing environment and actor gates protect access in this public repository. ## Linked Issues or Issue Description Related: #15705 and #15706. A search found no open PR that adds this workflow credential binding. **Describe the bug** Copilot Product and direct-protocol eval cells require `COPILOT_GITHUB_TOKEN`. The protected paid workflows do not currently pass that environment secret to the selected execution step. **Expected behavior** Only a selected Copilot cell receives this credential. Ordinary PR, build, catalog, image and reporting jobs receive none. **Steps to reproduce** Dispatch the trusted full-stack workflow for one Copilot cell. Credential validation fails before the task starts, although the protected environment can hold a valid token. **Environment** GitHub Actions in the public `paperclipai/paperclip` repository, with the reviewed workflow on `master` and the `runner-e2e-paid` environment. ## What Changed - Pass `COPILOT_GITHUB_TOKEN` in the paid Product and direct-protocol execution steps, only when `matrix.credentialName` selects it. - Add the missing Copilot protocol-catalog credential mapping and verify it reaches the selected execution matrix. - Add security checks for the selected-cell binding and bounded protected-job scope. Extend the scan that rejects provider credentials in unrelated workflows. - Document the Copilot credential scope. No secret values, account data or private test evidence are in this PR. ## Verification - Product workflow security: 16 tests passed on this head. - Protocol catalog and workflow security: 21 tests passed, including the Copilot matrix regression reproduced red before its fix. - Current-head required `ci / verify` and `ci / e2e` are green. Greptile is 5/5 with zero unresolved review threads. - The first CI attempt retained runner-shutdown failures; one failed-job retry passed on the unchanged head. Both attempts remain in [the CI run](https://github.com/paperclipai/paperclip/actions/runs/40342429106). - `git diff --check` passed. - A selected-credential scan found zero matching secret values in the seven changed files. - The environment setting was separately checked: `master` only, required approval from existing allowlisted code owners, and administrator bypass disabled. The new secret exists only at environment scope; no repository-wide duplicate was created. - The packaged Copilot runtime verified the selected token and the explicit model without a model prompt. A fresh local UI journey then created an agent, completed its first task and verified the downloaded artifact and cleanup. ## Risks The selected paid cells gain access to one additional environment credential. Missing credentials still fail closed. Default-branch checks, initiating/rerunning actor checks, protected environment access and per-cell credential selection remain in place. This PR adds no trigger and gives no credential to builds or ordinary PR checks. ## Model Used OpenAI Codex, based on GPT-6, with repository tools and browser regression tests. The exact runtime model identifier is not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either linked existing issues or described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links - [x] My branch name describes the change and contains no internal ticket id - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
49bd7a4ba6 |
feat(evals): retain portable Runner controller evidence (#17469)
Retain private Runner request evidence for offline cost reconciliation. Bound post-child reads, reject hidden paid tools, exclude configured credentials, bind incomplete results, and replace redacted proof atomically.
Verified at
|
||
|
DottaandPaperclip
|
a1f77821b9 |
perf(tasks): show recent conversation before historical activity (#17593)
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work.
> - A task page must show the instruction and current conversation
quickly.
> - The current page waits for related task data and historical run
transcripts.
> - These reads grow with task history, even when the reader only needs
the latest answer.
> - This pull request gives cold opens a small initial view and loads
older history on demand.
> - Messages and runs observed while the page stays open remain visible.
> - The benefit is a readable task before historical diagnostics finish
loading.
## Linked Issues or Issue Description
Related: #14667 and #16229. The first changes the reveal gate and scroll
anchors. The second adds general history pagination. This PR adds a core
task read and a folded recent-conversation view. It preserves the
default full-read API contracts. The shared files will need coordination
if either related PR lands first.
**What existing behavior does this improve?**
Cold opening an ordinary task with a long conversation or large run
logs.
**Current behavior**
Task enrichment and historical transcripts delay the readable
conversation. A cold open can fetch several megabytes of tool activity
before revealing saved messages.
**Proposed behavior**
Show the task instruction, recent saved replies, latest human input, and
the latest run's conversation data first. Fetch older messages and full
run activity when the reader chooses **More messages**. Retain new
messages while the page stays open.
## What Changed
- Add an authorized core issue view that skips related enrichment. Keep
enrichment on the existing detail read. Share task mutation, rollback
and invalidation updates with the core cache.
- Add a bounded recent-comment read. Preserve the first human
instruction when no description exists, including legacy rows with a
NULL author type. Keep company and task access checks and comment
redaction.
- Add optional recent activity and run-summary views. Fetch original
input receipts for visible comments separately so delivered messages
stay out of the queue. Reconcile watched unfinished runs after a
reconnect beyond the preview window, removing deleted runs after
reconciliation.
- Filter the latest native run's initial events to assistant messages,
request lifecycles and terminal results. Full event reads remain
available on expansion. Legacy fallback reads use a small tail extended
to the first complete record. HTTP cancellation aborts the storage read.
- Retain observed messages across cursor catch-up and new runs. Apply
local cancellation and deletion to both caches and their task aliases.
Refetch locally seeded history pages on expansion. Preserve expansion
across identifier and UUID links. Keep deep links, classic tasks and
Agent Chat on their existing full-history path.
- Update API documentation, design guidance, capability maps and
regression coverage.
## Verification
- Final acceptance on `9ea0b15ae44d9f6d2c9a9f824b4835f299f92979`: all CI
gates passed, including server/UI tests, all eight browser shards,
runner checks, full typecheck and build. [Final CI
run](https://github.com/paperclipai/paperclip/actions/runs/40326990328).
GitHub reports 55 successful checks, two intentionally skipped checks,
and one neutral Paperclip Review check described below. No checks are
pending or failed; Greptile is 5/5 and no review threads remain open.
- October 26 merge update: merged current `master`. The only conflict
was in task-page imports; both the compact-history hook and the new
live-update hooks remain. All CI tests and build jobs passed on
`044fd6733d`. Fresh review found canceled-comment cache drift, legacy
human-comment filtering, send-then-expand truncation, and partial legacy
log records. These defects were fixed in `af3d66a5bd`.
- Review-fix verification: 470 tests pass across ten affected suites,
including all six PostgreSQL cold-open tests without a helper override.
Cancellation/delete and send-then-expand regression tests failed before
the fixes and pass afterward. Coverage includes task identifier/UUID
reopen, stale in-flight reads, legacy NULL author types, local-file and
object-storage tails cut inside UTF-8 text, route forwarding, and
request cancellation. Full workspace typecheck and production build,
token gates, map validation and diff checks pass.
- Follow-up review found that a failed expanded legacy read lost its
record-alignment state on retry. `9ea0b15ae4` commits the expanded
window only after a successful read. The new regression fails before the
fix and passes afterward; all 19 transcript tests, UI typecheck, UI
build and token gates pass. Greptile reviewed this commit at 5/5 with no
actionable findings, and all review threads are resolved.
- The final Paperclip Review check ended neutral because its execution
did not complete; no passing Paperclip assessment is claimed. Its
earlier four concrete findings were fixed and regression-tested.
Greptile and the security scans completed successfully on the final
commit.
- The preceding `af3d66a5bd` CI run was canceled when superseded by the
retry fix. Its server shard 11 check briefly reported failure, but the
job API records cancellation and its test step passed all 1,750 tests.
This superseded run is not used as final CI acceptance.
- 549 focused tests pass on the final implementation across nine task
UI, activity-route, OpenAPI and PostgreSQL suites. These cover pending
history, desktop/mobile visibility, expansion, retained runs, message
bursts, cancellation recovery and queued-comment removal in both views.
- Legacy expansion keeps the latest answer visible while fetching and
deduplicating older output.
- Five PostgreSQL tests pass for core access denial, recent comments,
missing-description instructions, event paging and bounded activity/run
reads. macOS could not start another PostgreSQL cluster because it
exhausted its semaphore limit. These tests used a separate disposable
database on an existing local server through a temporary test setup
override. The helper was restored afterward.
- Full workspace typecheck and build pass with the final state fixes, as
do token gates and capability-map validation.
- Token gates, capability-map validation and `git diff --check` pass.
- Browser proof used a disposable task with 80 comments and six large
run histories. The cold view showed the instruction and recent exchange.
**More messages** revealed older comments. Navigation away and back
preserved expansion.
- Local HTTP sample: core 50 ms, recent comments 24 ms, recent run
summaries 17 ms. The latest native conversation response was 839 JSON
characters / 8 ms; its full event response was 1,099,005 / 86 ms.
Browser content paint was 1.34 seconds with development assets cached.
These are local observations, not staging latency claims.
- The first CI run passed every job except four board-send browser cases
and the review threshold. Those browser fixtures only mocked the old
comments endpoint; they now serve both recent and full reads, retaining
all message and attachment assertions. Both initial review findings were
fixed with regression tests. The next CI run passed every job. The
deeper review identified core-cache mutation, older delivery receipt and
retained run-status gaps; those are now fixed with regression coverage.
Final review is 5/5 with zero unresolved findings. Final-head CI
initially failed one unchanged runner shutdown test
(`runnerd-codex-transport.test.ts`, timeout-attachment case) with
`NativeSessionCloseUnrecoverableError`; all other CI jobs passed. The
same runner code passed on
|
||
|
DottaandPaperclip
|
41c42895fa |
feat(evals): enforce fixed Runner request allocations (#17452)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Runner evals execute real provider harnesses against maintained fixtures. > - Turn-level usage arrives after requests have already spent money. > - Parallel cells need distinct allowances before requests start. > - Missing output must preserve the remaining allowance and any known charges. > - This pull request adds a controller that reuses the existing request ledger and gateway. > - It makes a bounded local Runner attempt reviewable before hosted scheduling is enabled. ## Linked Issues or Issue Description Refs #17440. That PR added the explicit Runner gateway route. This PR owns request admission and retained campaign accounting for that route. The existing hosted matrix is not enabled or changed here. Its global allocation, rerun rejection, and private evidence archive remain separate integration work. ## What Changed - Validate a trusted manifest with exact request, CLI, package, daemon, model, and source identities. Assign each selected cell a distinct allowance. Count operating overhead once. Reject allocations above the campaign limit before launch. - Run native Codex or ACPX Claude through the shared request proxy and append-only ledger. Pass only the proxy credential to the child. Refuse local attempt reuse and preserve existing output files. - Close gateway admission on cancellation. Retain the ledger, receipts, route descriptor, child result, and cleanup evidence outside the disposable workspace. - Recompute campaign accounting from retained ledgers. Missing or invalid final evidence holds the full cell allowance while preserving known charges. Reports distinguish provisional and incomplete costs. - Accept namespaces containing only local function/custom tools. Disable web search in the controlled Codex profile. Continue to reject hosted tools, including tools hidden inside a namespace. - Add controller tests to the existing Runner static test path and document this source-checkout entry point. ## Verification - 20 controller tests pass. They use real child processes, local HTTP, and the actual ledger. Cases cover reservations before dispatch, concurrent cell isolation, caps, cancellation, replay rejection, lost artifacts, changed receipts, and Braintrust Claude routing. Upstream responses are synthetic. - 52 shared gateway tests and 55 Runner CLI tests pass. The namespace acceptance and web-search configuration regressions failed before the repair and pass afterward. - Runner TypeScript/Rust checks, Runner build, shared E2E typecheck, and `git diff --check` pass. - The first actual-harness probe found that Codex advertised both a local tool namespace and web search. It reached no upstream transport. That failure is retained; the compatibility repair does not allow paid hosted tools. - An earlier test failure showed that separate dynamic imports created different admission-error classes. The shared runtime bridge now loads one module graph. Another regression showed that a missing final report hid known ledger charges; the collector now preserves those charges. - The repaired actual-harness probe passed its control checks for native Codex and ACPX Claude: each reached exactly one synthetic upstream request after the ledger reservation; both retained the unresolved hold and route evidence; both recorded clean process exit. A scan of 41 retained files found no synthetic upstream credential. Deliberate provider rejection preserved the failed Codex turn and Claude timeout; these are not semantic passes. Prices, operating bounds, and the package archive identity in this probe were synthetic. - Review regressions reproduced lost request/receipt evidence, lost admission metadata, executable replacement after verification, and a SIGTERM-resistant descendant surviving the CLI. The repairs now pass. Executable bytes are staged before launch; a trusted Node resolution hook preserves relative imports. A symlink-path regression also failed before its repair and now passes. The final staged native probe repeated both one-request checks successfully. - Full current-head [CI](https://github.com/paperclipai/paperclip/actions/runs/40115932485) passed on `4781487215116afd10a415d4db53b1ec8e004dd5`: all 58 check runs and the legacy status are successful, neutral, or skipped. Greptile and Paperclip both reviewed this commit at 5/5. All five finding threads are fixed and resolved. No paid provider request ran during this change. ## Risks - This is a trusted source-checkout controller, not a published package command. It needs the repository's installed shared cost dependencies. - The manifest declares reviewed prices and full-lifecycle bounds. Synthetic tests do not qualify real prices, invoices, storage, or Braintrust metering. - The attempt-directory guard is local. Hosted fanout still needs unique global allocations, rerun rejection, and durable private retention. Do not enable the existing matrix on this change alone. - Forced child termination can leave cleanup unverified. That attempt remains incomplete and holds its full allocation. These checks do not claim a host security boundary against an adversarial target. - The current private publisher does not archive the separate controller directory. Native trace relay, private report integration, and live qualification remain open. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, tool use, and code execution. This runtime does not expose the exact model ID or context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Devin FoleyandPaperclip
|
1a3f604097 |
fix(db): honor the prepared-statement opt-in for ORM queries (#17673)
Drizzle calls Postgres.js unsafe queries, which disable preparation unless per-query options enable it. The existing client-level opt-in therefore does not prepare ORM queries. Honor explicit true across root and child clients. Preserve defaults, per-query options, lazy execution, and driver error handling. Document the pooler capability requirement. Test with 92 selected database and connection checks, workspace typechecks, server and database builds, and a read-only hosted ORM protocol probe. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
Devin FoleyandPaperclip
|
34409a92c4 |
fix(native-runtime): stop failover backups from carrying throwaway provider state (#17618)
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work
> - Native runner sessions on managed sandboxes keep a failover backup
of provider state, so a replacement sandbox can resume the session
> - Each backup copies the provider's runtime directory, with a
per-directory sha256 manifest, minus `excludeEntries`
> - Codex already excludes credentials and re-downloadable plugin
caches; OpenCode excluded nothing, and ACPX excluded only per-agent
credentials
> - OpenCode also leaked a temporary HOME on every launch, because
removing it failed on the read-only skill snapshot inside and the error
was swallowed
> - So every backup carried every earlier launch's HOME (with its npm
cache) plus re-installable plugin `node_modules`
> - This pull request removes those homes correctly, sweeps homes left
by earlier versions, and excludes the throwaway entries from OpenCode
and ACPX backups
> - The benefit is much smaller backups and live session directories,
while session history is still backed up and restored
## Linked Issues or Issue Description
Bug report described in-PR (bug report template).
**What happened?**
On one instance, current native failover backups totalled 24 GB across
about 300 sessions. For OpenCode sessions:
- `opencode/<session>/home-XXXXXX/`: 170 per-launch HOME directories
across 67 sessions, 9.3 GB. Each held `.npm` (about 111 MB of npm
cache), `.claude` (the skill snapshot) and `.local`.
- `opencode/<session>/config/opencode/node_modules`: 68 sessions, 4.3 GB
of OpenCode plugin installs.
One ACPX session also carried `acpx/<session>/data/pnpm/store` (229 MB
in the current backup, several GB in older copies).
Cause of the leaked homes: `startRuntime` in the OpenCode driver creates
a fresh `mkdtemp` HOME for each launch and seals the assigned skills
inside it (directories `0555`, files `0444`). `close()` and the
startup-failure paths removed it with `rm(..., { recursive: true, force:
true }).catch(() => undefined)`. That fails with `EACCES` on the sealed
directories, and the failure was ignored, so each launch left its HOME
behind.
**Expected behavior**
Per-launch homes are removed when the runtime closes, and backups hold
only state needed to resume a session. Caches that the provider
re-creates are left out.
**Steps to reproduce**
1. Run an OpenCode native session with at least one assigned skill on a
managed sandbox, over several turns or runs.
2. Inspect `<session>/` in the OpenCode runtime directory: one
`home-XXXXXX` per launch remains.
3. Inspect `failover-backups/current/opencode/<session>/`: the same
homes and `config/opencode/node_modules` are in the backup.
**Paperclip version or commit:** `master` at
|
||
|
DottaandPaperclip
|
aec86c794b |
fix(runner): preserve native work through controller outages (#17594)
Keep remote native work under its authenticated connection lease through controller outages. Hold durable ACPX tools through reconnection, align the native Claude HTTP MCP client budget, and wait for the recovered handler before dispatch. Preserve Stop, cancellation, exact identity and artifact checks. Verified with real-process regressions and a real Claude Opus 5.5 Daytona workflow: 125 seconds offline during useful work, original process and lease retained, no failed native tool calls, queued follow-up completed, independently verified ZIP and committed workspace receipts. All current-head CI gates pass and Greptile is 5/5 with all findings resolved. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
Devin FoleyandPaperclip
|
efa2724f9a |
fix(db): pace bounded response capture output (#17564)
Close database response measurement at its existing deadline, then emit the bounded metadata at 64 records per second. Preserve group indices, counts, privacy, and query/transaction outcomes. Stop the unreferenced timer when the output queue is empty. Validation: 32 capture and real PostgreSQL tests, full build and typecheck, exact-head Greptile 5/5 (medium), full green CI, and no unresolved comments. A 2,200-query fixture reconciles all 2,048 groups and its summary; dedicated-pool coverage awaits deferred output. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
Devin FoleyandPaperclip
|
a12329b8a9 |
fix(db): complete bounded response attribution (#17537)
Raise bounded database-response attribution to 2048 groups and add numbered 128-entry log pages. Preserve the overflow bucket and exact completed-call and byte totals. Document complete page retrieval and validation. Verified with 30 focused tests, the full build, and typecheck. All CI checks pass. Exact-head Greptile is 5/5 at Medium impact with no unresolved comments. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
Devin FoleyandPaperclip
|
71f06d84dc |
fix(db): keep bounded response captures attributable (#17522)
Keep more caller groups in one bounded database response capture and verify public caller paths in the production bundle. All local checks and CI passed; Greptile scored the exact head 5/5 with no unresolved comments. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
d0c8f0b090 |
fix: explain project access denials and add an authorized repair path (#17509)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Low-trust agents may read only the projects in their allowed scope. > - Workspace setup correctly stops a task outside that scope. > - The task previously showed a generic failure, and agent settings did not offer an additive repair. > - This pull request explains the denial and lets authorized operators add projects while preserving other restrictions. > - The saved task then offers an explicit retry. Saving access does not start work. ## Linked Issues or Issue Description **What happened?** A project access denial appeared as an ordinary failed attempt. Operators could not add a second allowed project from agent settings. **Expected behavior** The task should explain the denied access and link to a working permission editor. The editor should preserve existing scope and restrictions, enforce operator authority, and retain drafts after a failed save. **Steps to reproduce** 1. Assign a low-trust agent to a task whose project is outside its allowed projects. 2. Let workspace setup deny project access. 3. Open the task, follow Manage project access, add the project, save, reload, and return to the task. 4. Confirm that the saved message remains and Retry task is a separate action. ## What Changed - Record a typed, non-retryable project denial before provider dispatch. Project denials do not enter automatic continuation or provider comment policy. - Project current and historical setup denials into a safe task notice. Keep unrelated failures, prior assignees, and approval/pause/budget gates separate. Recheck current project authority, including standard trust and root-issue grants. Accepted retries keep their working state. - Add board-only project-access reads and additive saves under the agent row lock. Enforce company scope, agent-management authority, pending-approval restrictions, and an audit record for new additions. - Add the production notice and searchable multi-project editor, including read-only, saving, cancellation, and failed-save states. Preserve the return-to-task query through canonical agent redirects. - Register board-only project-access routes in OpenAPI with a strict request schema. - Add production Storybook examples and API/persistence/concurrency regressions. Retain the approved design snapshot and update the behavior maps and implementation contract. ## Verification - Passed: `pnpm -r typecheck`, `pnpm build`, `pnpm build-storybook`, `pnpm check:token-gates`, `node scripts/check-maps.mjs`, and `node scripts/storybook-catalog.mjs`. - Passed: 11 permission/projection tests and 30 run-completion tests against an isolated PostgreSQL database. These cover historical failures, concurrent additions, idempotent audit, preserved restrictions, foreign projects, read-only users, agent self-grants, and no automatic wakeup. - Passed: all 27 OpenAPI route tests, including route coverage and the new board-only request/response contract. - Browser: the real app displayed a historical denial; Manage project access opened the editor; saving retained the original project and survived reload; Return to task showed the ready notice. Saving created one audit event and zero runs. Explicit Retry returned 202 and reached the separate sandbox requirement, which remained enforced. No external provider was used. - All five production Storybook interaction checks pass: blocked task, save, cancellation, failed-save recovery, and read-only settings. These use fixture callbacks; durable authorization is covered separately above. - `pnpm test:run` was attempted locally but stopped after repeated embedded PostgreSQL startup failures caused by macOS semaphore exhaustion. Full-suite CI is required before merge. The separate UI suite passed all 9,095 tests. The 41 focused server regressions also passed after rebasing onto the latest master. A further 26 recovery-classification tests pass, including a red-to-green regression for automatic continuation after a project denial. ## Risks - No schema migration. The new endpoint only adds same-company project IDs and preserves the rest of the stored policy. - The historical notice recognizes one exact setup denial message. It does not display arbitrary run errors or disclose a different source project's identity. - The retry still passes all current admission and isolation checks. The notice does not promise that other execution requirements are satisfied. - Related open work: #13309 covers inherited secret restrictions; #11234 covers the broader repository catalog. This change does not replace either policy surface. ## Model Used OpenAI Codex (GPT-6), with code execution and browser tools. The precise runtime model ID and context-window size were not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass (focused server and full UI suites; full-suite infrastructure limit above) - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Devin FoleyandPaperclip
|
a3675ddcdf |
Add bounded database response size diagnostics (#17504)
Count database response value sizes by public caller and static table for one explicit UTC window of at most 60 seconds. Keep diagnostics off by default and exclude SQL, parameters, returned values, errors, and credentials from output. Preserve lazy execution, transaction boundaries, and no-replay behavior. Verification: full build and recursive typecheck; 63 focused database and PostgreSQL tests; eight physical disconnect tests with capture enabled. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
044628f59e |
Add explicit native tracing to the Runner eval CLI (#17483)
## Thinking Path > - Paperclip manages AI agents and their work. > - Runner Evals exercise real providers against a mock control plane. > - The controlled gateway route creates a disposable provider home. > - That route had no explicit native trace configuration. > - Codex skips normal home preparation for a preconfigured home. > - This change validates and applies the existing observability launch contract before provider creation. > - Operators can collect native evidence without passing vendor keys to the child. ## Linked Issues or Issue Description Refs #17440. Related: #17452 and #17469. This PR is based directly on master and does not include those pending controller changes. **Current behavior** The eval-session CLI cannot pass an explicit native observability launch. Passing environment alone cannot configure Codex in the disposable gateway home. **Proposed behavior** Accept an opt-in descriptor bound to the gateway attempt. Validate the existing provider launch contract, add a scoped relay token in memory, and apply it only to the owned provider home and environment. **Reason and benefit** A controller can route native evidence through its bounded local relay while the CLI keeps configuration distinct from verified ingestion. This is the CLI contract; hosted controller integration remains separate work. ## What Changed - Add a strict, bounded descriptor reader for native Codex and ACPX Claude. Bind campaign, execution, attempt, project, trace, span, and mock authority identifiers. - Reuse Runner observability materialization and redaction. Preserve the model gateway route. Reject upstream destinations, extra environment fields, credentials in headers, and unknown exporter settings. - Add `--observability-config`, requiring an explicit gateway. Keep direct and unconfigured paths unchanged. - Retain secret-free configured/unverified metadata in success and failure artifacts. - Document the contract and controller responsibilities. No workflow, runtime dependency, paid schedule, or public report changes. ## Verification - Eval CLI regression selection: 81 passed before the test-only relocation. The 17 new cross-package observability tests now pass from `tests/runner-e2e/runner-eval-observability.test.ts`; their integration-suite typecheck also passes. They use the actual shared launch generator, real disposable homes, CLI success/failure paths, and credential redaction. The first run exposed two test setup references introduced by a broad edit; both were corrected and its failure log remains retained. - Runner TypeScript build and release daemon build passed. Full `pnpm -r typecheck` passed. - Actual native Codex and ACPX Claude reached a local receiver with authenticated OTLP protobuf and the expected Braintrust parent. Codex sent four log and two trace batches; Claude sent three log and one trace batch. All recorded processes exited, disposable homes were removed, and 29 retained files contained neither synthetic credential. - The local model endpoint deliberately returned 401. Codex retained a failed terminal turn; Claude retained a provider timeout. These are route/export proofs, not semantic eval passes. Native retries reached only the fake endpoint. No paid provider or Braintrust request was made. - The first local probe failed before provider launch because its TypeScript loader path was wrong. It was corrected to the installed loader; both attempts remain retained. - Full repository build passed at initial head |
||
|
DottaandPaperclip
|
0eb2c964a7 |
fix(agents): allow runner qualification with agent configuration permission (#17448)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Agent owners configure the harness and computer used for that work. > - A legacy runner upgrade must first qualify an unverified sandbox. > - The upgrade called an instance-admin probe, so a company owner with agent configuration permission received HTTP 403. > - This pull request qualifies the saved computer through the agent's existing configuration authorization. > - Company owners can complete the upgrade while general instance environment management stays restricted. ## Linked Issues or Issue Description Refs #17361. Related: #17122. **What happened?** Upgrade runner displayed for an existing Claude agent on a Daytona Paperclip Computer. Clicking it returned “Instance admin access required” for a cloud company owner. The agent remained on the legacy runner. **Expected behavior** A caller permitted to update the agent can qualify that agent's saved computer and upgrade. Instance-admin access is not required for this agent-scoped operation. **Steps to reproduce** 1. Sign in as a cloud company owner with `agents:configure` and no instance-admin role. 2. Open an existing legacy Claude agent on an unqualified Daytona snapshot. 3. Select Upgrade runner. ## What Changed - Add `POST /api/agents/{id}/qualify-runner` with the existing `agent_config:update` guard. - Reuse the server's effective environment resolver. The submitted environment ID only detects stale state; it cannot select a different computer. - Preserve company binding checks, check binding again at lease acquisition, and record agent-scoped audit activity. - Return qualification status and the lease receipt without instance-level provider metadata or raw provider errors. - Use the new endpoint for the existing-agent upgrade. Keep the generic environment probe restricted to instance admins. - Disable Upgrade runner when an unsaved computer change resolves to a different target, with a save-first explanation. Other draft settings still accompany an upgrade. - Add real-database authorization regression tests and update the API contract and access maps. ## Verification - Red: the existing UI called the instance-wide probe and failed the new agent-scoped call assertions in eight upgrade cases. - Green: 349 related route, provider qualification, runner selection, and UI tests passed. - Green: 12 authorization cases passed with real PostgreSQL, real HTTP routes, real saved role grants, and a simulated provider boot. They cover the cloud owner, removed grants, foreign agent/computer, inherited targets, stale target, invalid body, archived/local target, redacted failure, audit scope, and unchanged agent settings. - Review regression: two cases first reproduced the enabled upgrade button after an unsaved computer change; both now pass, along with all 168 related UI tests. Qualified/unqualified computers and restoring the saved selection are covered. - All 26 OpenAPI cases passed; the 12 real-database authorization cases also passed after rebase. - The host exhausted PostgreSQL semaphores. For the 12 cases, a temporary local-only setup used disposable PostgreSQL 18 in Docker instead of the embedded database launcher. The committed tests use the normal embedded database helper. No authorization decisions were mocked. - `pnpm -r typecheck`, `pnpm build`, `pnpm check:token-gates`, map validation, module-boundary validation, and `git diff --check` passed. - `pnpm test:run` was attempted but did not establish a full local pass: the host could not start embedded PostgreSQL clusters. All current-head CI gates passed, including the full suite in CI. The commitperclip review workflow timed out before reporting any executed steps on its first attempt and passed on one retry. - Fresh staging acceptance remains pending deployment. Verify persisted native configuration and a successful runtime test on the existing Daytona computer after merge. Earlier live Daytona proof from #17361 is separate evidence. ## Risks - Qualification can start a fresh, temporary sandbox. It uses the existing probe lifecycle and cleanup contract. - The new path must remain scoped to the saved agent and its company. Denial and redaction tests cover that boundary. General instance probe authorization is unchanged. - No migration, role change, credential change, or automatic agent upgrade is included. ## Model Used - OpenAI GPT-6 through Codex, with reasoning, code execution, and browser tools. The exact deployed model variant and context-window size were not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
7ae81c9fa7 |
fix(slack): support rich questions and complete review delivery (#17255)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Questions and review decisions must work in the conversation where a task started. > - Slack omitted question context and did not publish several review kinds. > - Some answers saved in Paperclip also lost their authorized continuation or their final Slack reply. > - A settled card and a successful run did not prove that the user received the result. > - This pull request preserves complete questions and follows the existing answer and delivery receipts through continuation. > - Users can read the request, submit an authorized decision, and receive the resumed task's result in the same conversation. ## Linked Issues or Issue Description Refs #17045 and #17218. **What happened?** Slack question cards could omit instructions and editable starting text. Checkbox reviews, item verdicts, task suggestions, and connection requests had no publication path. Complex question answers submitted in Paperclip could fail continuation authorization. Several review continuations completed without delivering their result to Slack. **Expected behavior** Show the actual question and context. Use native Slack controls when they preserve the form contract. Otherwise provide an explicit Paperclip review link. Persist one authorized response, resume through the existing delivery path, settle the card, and deliver the useful result to its originating conversation. **Steps to reproduce** 1. Start a task in Slack and ask a text or choice question with instructions and help text. 2. Submit native text and choice responses, then submit a complex form through Paperclip. 3. Accept a checklist, submit item decisions, select suggested tasks, or decline a connection request. 4. Inspect the saved answer, continuation run, original Slack card, and final Slack reply independently. ## What Changed - Preserve full prompts, context, descriptions, recommendations, help, and editable starting text. Use direct buttons for short choices, inline searchable selects for longer lists, and native checkboxes with explicit submission for multiple choices. - Add native custom-answer and validated-text modals, ordinary confirmation buttons, checklist decisions, per-item verdicts with reasons, suggested-task creation/dismissal, and connection decline. Keep governed approvals, permission grants, and forms that cannot be represented faithfully in the full Paperclip review surface. - Keep Slack-specific canonical-validation feedback in its owning platform module. Shared callbacks delegate without interpreting provider presentation. - Reuse the existing canonical interaction transactions, callback receipts, publication queue, and continuation delivery. Share ordinary confirmation handling across Telegram and Slack. Recheck current authorization and commit response receipts atomically. - Settle cards with the saved answer and remove stale controls. Ignore passive Slack link callbacks so opening a task does not display a false answer error. - Validate exact multi-select receipt mappings during native continuation. Recover automation-origin connection replies from their published prompt after validating the durable connection receipt. - Keep oversized multi-select handoffs free of unused native tokens, and retain the complete Paperclip form for 101–128 choices plus Other. Make the original-conversation audience explicit before submitting a Slack modal. - Add rollback, provider serialization, callback, continuation, and negative authorization regressions. Update the connector playbook and feature/access maps. No React or application UI changes are included. - Add a durable Q01–Q11 qualification index and provider coverage matrix. Require distinct functional, good-experience, and best-available-control assessments; real unanswered/answering/settled evidence; saved results, continuation and navigation proof; and retained failures. Mark other providers unaudited rather than assuming Slack parity. ## Verification Candidate: `e0017d446b1016a38d65382366cd1408cce33039`. The eleven-core-format live Slack gallery is pinned to `c3d458b2a06b660448940fd8cf33c5906f0058d5`; subsequent boundary fixes and the sharing notice have automated verification, not new live screenshots. - Real Slack UI testing covered all 11 formats: text, closed choice, long single select, multiple choices, choices with a custom answer, pattern-validated text, confirmation, checklist, item verdicts, suggested tasks, and connection requests. Each has unanswered/answering/answered screenshots as applicable, an inspected saved result, one continuation, and an individual experience review in the local HTML gallery. - The dedicated private QA app/channel used signed callbacks to the local candidate. A separately approved password-protected, read-only preview proved connection-review, suggested-task review, and created-task navigation. Anonymous access, writes, and credential routes were denied. No OAuth grant was attempted. - Real testing exposed and repaired a multi-select continuation failure and a misleading error on link-button clicks. Failed attempts remain in the evidence. The final navigation replay opened the exact pending connection request and declined it in Slack without granting access. Suggested-task creation produced the selected child with the correct parent and description. - Final review-fix verification: **32 form tests**, **28 provider/form-rendering tests**, **30 publication tests**, and **171 native continuation/authorization tests** passed. Earlier **3 atomic-resolution rollback**, **9 targeted callback/publication/Telegram**, and **28 review-continuation** cases passed. The final Slack/Teams modal subset also passed. The automation-origin connection cases, 101/128 custom-choice cases, and oversized multi-select continuation reproduced before their fixes and passed afterward; failing logs remain retained. These counts describe relevant runs, not a reliability percentage. - CI on `b5885b85fd` exposed a test fixture that assumed every Slack modal block was an input; the new sharing notice correctly renders a non-input block. The fixture now selects input blocks safely. The exact signed-modal correction/one-continuation test passed locally, together with 14 Slack form/modal tests. The failing CI log and its specific cause remain recorded. - `pnpm -r typecheck` and `pnpm build` passed. Server typechecking was repeated after the final source changes. Map and module-boundary checks and `git diff --check` passed. - The full `pnpm test:run` was attempted and stopped after repeated embedded PostgreSQL semaphore exhaustion on this host. Its 1,395-test chat suite found two stale native-control assertions; both were corrected and re-tested in a focused run. Focused database suites used separate disposable databases on existing local PostgreSQL. The full suite is not claimed green; new-head CI must complete before merge. - Final-head CI passed on `e0017d446b1016a38d65382366cd1408cce33039`: all test/build jobs and required `ci / verify` and `ci / e2e` gates are green ([run](https://github.com/paperclipai/paperclip/actions/runs/40115430194)). Greptile and Ponytail both reviewed this exact head at 5/5, the security scan passed, and all review threads are resolved. - All three QA app callbacks were restored to staging. Both temporary tunnels and local proxies were closed. The local endpoint was archived and its local credential copies removed. The candidate app was stopped. Limits: provider execution used a deterministic Claude protocol fixture, not an LLM. This is local real-provider UX proof, not deployed staging or broader Codie/Claude qualification. The public preview intentionally blocked ancillary artifact routes, so its destination pages could show a partial-history warning. The complete permission-grant flow and every oversized/mixed-form branch were not driven manually. Earlier fixture, test, and provider failures remain recorded. ## Risks - New reply routes require durable origin proof. Incorrect receipt matching could lose a reply or route it incorrectly. Tests cover missing, forged, mismatched, and revoked authority. - Governed forms keep their Paperclip resolution checks. This change does not grant new approval authority to Slack. - Existing completed review results can become deliverable through the repaired publication path without another provider run. - No schema, lockfile, workflow, public task-status model, or runtime-instruction change is included. This branch is not deployed. ## Model Used OpenAI Codex, GPT-6 family, with code execution and tool use. The exact backend model ID and context-window size are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either linked existing issues or described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links - [x] My branch name describes the change and contains no internal Paperclip ticket id - [x] I have run focused tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green on the new head - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups on the new head - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
d1fdfced9d |
feat(evals): route Runner sessions through a controller gateway (#17440)
## Thinking Path > - Paperclip manages AI agents and their work. > - Runner Evals test provider behavior against a mock control plane. > - A campaign budget needs to observe and admit each provider request. > - The existing eval CLI does not select the provider-home mode needed for a controller-assigned route. > - A proxy URL alone therefore cannot establish controlled Codex routing. > - This change adds an explicit route with an isolated home and retained identity. > - The controller can use that route as the basis for request accounting and admission. ## Linked Issues or Issue Description **What happened?** The direct Runner eval CLI passes the ambient environment to the native transport. The transport deliberately removes an environment-only preconfigured-home marker unless its explicit option is also set. This prevents a controller from assigning a Codex gateway through that marker alone. **Expected behavior** A controller can explicitly route a supported eval attempt through its local proxy. The provider receives the proxy token and isolated configuration. The artifact identifies the assigned route on success and failure. Ordinary direct runs keep their current path. **Steps to reproduce** Run the new CLI tests for `--gateway-config`. They validate the descriptor, inspect the actual transport inputs and provider sanitizers, and check retained artifacts and owned-home cleanup. The local probe also ran built native Codex and ACPX Claude against a fake loopback API using only synthetic credentials. Related request-accounting work: https://github.com/paperclipai/paperclip/pull/17432. ## What Changed - Add an explicit gateway descriptor bound to the requested provider, exact model, campaign, execution, and attempt. - Accept only the controller's IPv4 loopback URL and a separate proxy token. - Create a fresh provider home for native Codex and ACPX Claude. Exclude ambient provider credentials, subscription logins, proxies, and startup hooks. - Retain route identity and descriptor hash in successful and failed artifacts. Remove the owned home in finalization. - Document the CLI contract and its limits. The Runner eval CLI owns this behavior. It uses the existing native preconfigured-home contract. Production routing, permissions, and provider protocols are unchanged. ## Verification - Initial verification passed 56 focused gateway/CLI tests and 246 CLI and adjacent provider-isolation tests. After the admission repair, all 247 CLI and adjacent provider-isolation tests passed, with TypeScript checks and build passing again. - Runner TypeScript checks and build passed. - The native daemon rebuilt from the candidate base with locked, offline dependencies. - Local fake-API probe: six authenticated Responses requests from native Codex and six authenticated Messages requests from ACPX Claude reached the assigned endpoint. Claude also made an unauthenticated hello check. The final commit retained both route artifacts. All 27 retained files were scanned for the synthetic proxy token and contained none. Both temporary homes were removed, and all four recorded provider/daemon PIDs were absent after exit. - The fake API intentionally rejected inference. Codex retained a failed turn; Claude retained a timeout. These are route observations, not passing semantic evals or live provider qualification. No paid provider calls ran. - The first focused command used the wrong directory and discovered no tests. It was corrected before the successful runs above. - Full Runner validation: 3,967 tests passed, 13 skipped, and five failed in three unchanged installation and native-settlement files. The failures were two installation timeouts and three retained-cleanup settlement failures. One isolated run passed all 109 applicable tests with six skips. This shows recovery, not a proven cause for the original five failures; those remain retained. Hosted [CI 40111220754](https://github.com/paperclipai/paperclip/actions/runs/40111220754) passed on final commit `a5c47b64e56c1228fca718b72a81ddc3b1d185a4`, including repository typecheck, build, tests and Runner checks. All 57 checks passed or were neutral/skipped, and the legacy status passed. Greptile and Paperclip reviews both scored 5/5 on that commit; no findings remain unresolved. - Rebased onto current master `71c04c0bfb205cc9e7f7bc20c29fd76051b44cc4`. Its added change concerns server cleanup; Runner source is unchanged. - An actual ACPX Codex probe failed before reaching the fake API. Review and source inspection confirmed that Runner deliberately rejects its preconfigured home to protect Runner state. This route is now rejected before service creation. Both explicit and default ACPX Codex admission regressions failed before the repair. The unsupported lane is not presented as qualified. ## Risks This is an explicit new CLI route. Unsupported providers fail before service creation. Upstream route is a controller declaration, not independent billing evidence. The local probe saw six Codex requests where legacy usage counted one, so the controller's request ledger must remain the budget source. Campaign allocation, proxy lifecycle, hosted workflow wiring, dedicated OTLP relay routing, and paid qualification are separate remaining integration steps. This PR enables none of those schedules. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, tool use, and code execution. The exact serving model ID and context-window size are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used, including the information available in this session - [x] I checked ROADMAP.md; this is eval infrastructure, not a competing core feature - [x] I searched for duplicate PRs for this exact route change - [x] I described the issue in-PR using the bug template labels - [x] I have not referenced internal or instance-local issues - [x] My branch describes the change and contains no internal issue ID - [x] Relevant focused tests pass - [x] I added tests for the new contract and lifecycle - [x] I updated the relevant documentation - [x] I described the risks and unqualified behavior - [x] All current-head Paperclip CI gates are green - [x] Current-head Greptile is 5/5 with no unresolved findings - [x] I will address review findings before merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
87bf89756e |
fix: support Daytona runner upgrades and ignore obsolete Claude options (#17361)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Existing Claude and Codex agents can upgrade to Paperclip Runner. > - A Daytona computer needs platform qualification before Runner becomes its default. > - The upgrade notice required that default, so it hid the action that could start qualification. > - Saved Claude turn limits also caused conversion to fail even though Runner can run without them. > - This pull request verifies the computer on click and ignores obsolete Claude options with a warning. > - Agents retain their model, account, environment, instructions, and permission policy. ## Linked Issues or Issue Description Related: #17122. **What happened?** A legacy Claude agent on a Daytona snapshot had no upgrade button until its environment had platform qualification. A saved `maxTurnsPerRun: 1000` then caused conversion to return HTTP 422. **Expected behavior** The upgrade action verifies a supported Daytona computer and converts the agent. Unsupported tuning options produce a visible warning and do not block execution. **Steps to reproduce** 1. Create a legacy Claude agent with `maxTurnsPerRun: 1000` on an active Daytona snapshot. 2. Open Runtime before the environment has platform qualification. 3. Try to upgrade to Paperclip Runner. **Paperclip version or commit** After #17122, reproduced against `9c442cf48e`. ## What Changed - Offer upgrade on probe-capable Daytona snapshots and immutable sandbox images. Perform the public environment probe only after a click. Save after fresh adapter discovery confirms support. - Resolve the actual computer for explicit, instance-default, and hidden managed environments. Preserve pending edits and reject stale probe results. - Ignore Claude `maxTurnsPerRun`, `chrome`, `warmHandleIdleMs`, and `acpWarmHandleIdleMs`. Return structured warnings from create, hire, and update. Show warnings in Runtime and after a save. - Keep validation for commands, credentials, filesystem, network, and permission settings. - Retain the legacy turn limit in the existing live conversion fixture. Update regression tests, upgrade stories, and capability maps. ## Verification - Red: the old route returns HTTP 422 for the saved turn limit; the old form hides the Daytona upgrade notice. - Green: 417 focused server and form tests pass. A further 188 tests pass after the review fixes. Qualification coverage includes failed probes, unsupported targets, changed settings, managed defaults, unsaved environment variables, and forced Kubernetes targets. - `pnpm -r typecheck` and `pnpm build` pass. Token gates, map checks, map tests, and Storybook catalog checks pass. - Real local Claude Opus 5.5 conversion and task pass, with the downloaded artifact checked against seeded source bytes, SHA-256, and Git commit. - Real browser upgrade and full Claude Opus 5.5 task on a disposable named Daytona snapshot pass at `b81cec7d8c`. The saved agent retains its model, high effort, credentials, timeout, and grace settings while dropping the turn limit with a visible warning. Exactly one successful run completed the task. The downloaded artifact matched independently seeded source bytes, SHA-256, and Git commit; the source was unchanged, no interactions remained pending, and the delivery link survived a browser reload. The snapshot used current runtime assets built from public source. The supplied Daytona account could not access the original staging snapshot, so this verifies an equivalent local fixture rather than that exact snapshot. Disposable test resources were cleaned up. - Earlier live attempts remain recorded: an image-based sandbox creation failed, and the first named-snapshot probe returned failure with unresolved provisional cleanup identity. Later named-snapshot probing and the task succeeded; the initial creation failures are not attributed to this PR or declared resolved. A separate attempt using the Oct 18 cached runtime failed the current provider-pack compatibility check; rebuilding the runtime from matching public source removed that mismatch. Owned resources were identified and cleaned up, with diagnostic databases retained. - 54 checks passed at `b81cec7d8c`, with no failed checks; Greptile and Paperclip Review both scored 5/5, with both review threads resolved. Master advanced during the live proof, so the branch was rebased with both sides of additive import/test conflicts retained. On the rebased head `2b74222b94`, 472 focused tests, UI typecheck, and UI build pass. All required CI gates pass; Greptile and Paperclip Review both score 5/5 on this head, with no unresolved review threads. The PR is mergeable. The serial local `pnpm test:run` was stopped after the CI test lanes passed; that local run is not claimed as a pass. ## Risks - An explicit upgrade can allocate a temporary probe sandbox and take several minutes. Probe failures leave the agent on its existing runner. - Ignored options no longer enforce a turn cap, Chrome integration, or legacy idle-time tuning. The UI warns about each active ignored option. - Unsafe or ambiguous execution settings still fail validation. This change does not weaken environment qualification or runtime artifact checks. ## Model Used OpenAI Codex, GPT-6, with reasoning, code execution, and browser tools. The exact deployment model ID and context-window size are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Nicky Leach
|
aa0037a137 | Support Codex warm plugin task reuse and recovery (#17390) | ||
|
Devin FoleyandPaperclip
|
724fa6144d |
fix(opencode-local): preserve the agent during sandbox tool OOMs (#17398)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - The Legacy OpenCode adapter runs an agent and its shell tools in the same sandbox. > - A large build can exceed the sandbox memory limit and trigger a kernel OOM kill. > - The kernel can kill OpenCode before its tool, which loses the active turn. > - OpenCode provides a shell environment hook that can change only the tool subprocess. > - This pull request uses that hook to prefer Bash tool processes as OOM victims. > - The benefit is that OpenCode can report a failed tool and continue the task. ## Linked Issues or Issue Description **What happened?** An OpenCode sandbox run ended with exit 137 after a TypeScript process had also been killed. Its cgroup limit was 4 GiB. It had completed useful tool work and had not reached its execution deadline. The original sandbox was deleted, so its historical OOM counters are unavailable. Memory pressure is the leading explanation, not a proven historical kernel diagnosis. **Expected behavior** When a shell tool exhausts sandbox memory, the kernel should prefer that tool over OpenCode where possible. The adapter must preserve configured shell startup, model, permissions and resource limits. **Steps to reproduce** Use a disposable 4 GiB Linux sandbox with OpenCode. Retain a 1,200 MiB buffer in the agent and run a Bash tool with fourteen 250 MiB workers. Compare the same workload before and after installing the generated hook. Record the cgroup memory events, agent exit and final response. Refs: #16844. That PR covers the same protection for Claude. This PR adds the OpenCode-owned hook. Refs: #17254 for the separate shutdown recovery path. This PR does not change retry policy. ## What Changed - Generate a dependency-free OpenCode `shell.env` plugin and a Bash startup file in the isolated runtime config. - Stage those assets only for sandbox execution. Resolve the startup file relative to the copied plugin. - Set child `oom_score_adj` to `1000` on Linux. Keep the parent environment unchanged and honor existing `BASH_ENV` values, including an empty value. - Keep command execution available if the optional plugin cannot be installed or the process priority cannot be changed. - Materialize linked plugin directories inside the isolated config before writing. Keep the operator directory and its plugins unchanged. - Cover relocation, shell output and exit status, operator settings, local execution, remote staging and failed installation. Update the adapter documentation and verification recipe. ## Verification - Red before the fix: the new runtime regression failed because the sandbox shell plugin was absent. The existing 15 runtime config tests passed. The same regression passes with the fix. - Apex first found a host-write risk through linked plugins directories (4/5). A new regression reproduced the host directory modification, then passed after the installer materialized the copied link. The finding is resolved. Apex reviewed current head `df82a8f51725a197c0ea3ebeb942488509f32b34` at 5/5 with no actionable issues remaining. - The full OpenCode adapter suite passes: 87 source tests. The adapter typecheck and map checks pass. - Real sandbox acceptance used the plugin generated from `df82a8f51725a197c0ea3ebeb942488509f32b34` and OpenCode 1.18.34 in a disposable 4 GiB Daytona sandbox. A controlled local model endpoint issued an actual Bash tool call; no real model credentials were used. The tool reported priority `1000`; the parent reported `0`. - Controlled OOM comparison: without the guard, OpenCode was killed by SIGKILL after 8.95 seconds and the cgroup recorded one OOM kill. With the guard, five workers were killed, OpenCode exited `0` after 21.87 seconds, and its final response was `fixture-complete`. The cgroup counter increased by five. Both attempts used the same allocation and workload. The fixture started with a linked operator plugins directory. The installer preserved that directory, staged a real plugins directory, and passed both the process-priority probe and OOM comparison. The fixture allocation was deleted afterward. - The first live probe was not accepted because the harness could not parse the provider response framing. After correcting the decoder, the unchanged plugin passed the probe and the controlled OOM comparison. Both initial disposable allocations were deleted. After the review fix, a third allocation repeated the probe and OOM comparison with a linked plugins directory. That allocation was also deleted. The initial comparison had the same agent-killed/agent-survived result (8.42 seconds without the guard; 19.42 seconds with it). - Repository `pnpm -r typecheck` and `pnpm build` pass. The full CI matrix passes on `df82a8f51725a197c0ea3ebeb942488509f32b34`, including tests, browser shards, Runner checks, build, policy and Apex. [CI run](https://github.com/paperclipai/paperclip/actions/runs/40043947375). - The first full local test attempt had 15 embedded PostgreSQL startup failures. Direct `initdb --version` and `postgres --version` probes showed missing bundled library symlinks and SIGABRT. Running the bundled symlink hydration script restored both executables. All 15 affected cases pass after that setup repair. No product code changed for it. - Two existing workspace tests assumed that new Git repositories have no `main` branch. They failed under the host default branch setting of `main`, then passed with a per-command `init.defaultBranch=master` setting. The global setting was preserved. The full local test command was restarted with hydrated libraries and that controlled fixture default. It had no reported assertion failures when it was stopped after the complete CI matrix passed. It did not complete locally; full-suite verification comes from CI. The first failed invocation was also stopped. - The original tenant task has not been retried or recovered by this PR. Its board API credential is expired. The original hosted workflow and the separate shutdown recovery remain outside this acceptance result. ## Risks - This is a preference, not a memory reservation. It cannot guarantee that OpenCode survives every OOM condition, including memory growth inside OpenCode itself. - The guard applies to non-interactive Bash tools. Other shells ignore `BASH_ENV`. Existing operator startup settings take precedence. - Older OpenCode releases without the shell environment hook will not apply the preference. No model, permission, deadline, memory quota or retry limit changes. - No schema change or migration. ## Model Used OpenAI GPT-6 through Codex, with repository tools and code execution. The exact deployment variant and context window are not exposed by this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes: #` / `Refs: #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
|
cc127f5516 |
fix(runner): drop unusable musl builds and duplicate Copilot from provider packs (#17387)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - The runner release builds a Linux "provider pack". The pack is a self-contained runtime with Node, Claude Code, Codex, and OpenCode. The Daytona runner image and SSH hosts use it. > - The release artifacts for this pack are large: 1,182 MB for linux-x64 and 543 MB for linux-arm64. > - Most of that size is necessary. But about 20–30% of it is code that the pack can never run. > - pnpm 9 installs all libc variants of optional platform packages. Thus the pack includes the musl (Alpine) builds of Claude Code and OpenCode. The pack runs only its copied glibc Node, so these builds are never used. > - On linux-x64, the npm Copilot package is also a byte-identical copy of the verified Copilot provider asset (162 MB). > - This pull request removes these unused bytes from the pack at build time. > - The benefit is smaller release artifacts and Docker images, and faster uploads and downloads. ## Linked Issues or Issue Description **What happened?** The runner release workflow uploads `runner-provider-packs-linux-x64` (1,182 MB) and `runner-provider-packs-linux-arm64` (543 MB). The darwin daemon artifacts are about 10 MB. The darwin artifacts are a different artifact type, so the comparison is not direct. But inspection of the extracted Linux packs found real waste: - `@anthropic-ai/claude-agent-sdk-linux-*-musl` (about 225 MB uncompressed) - `opencode-linux-*-musl` and `opencode-linux-x64-baseline-musl` (about 180 MB each) - on linux-x64, `node_modules/.pnpm/@github+copilot-linux-x64@1.0.88/.../copilot`, which has the same SHA-256 (`0059754c…`) as `provider-assets/copilot/linux-x64/copilot` **Expected behavior** The pack contains only code that its copied Node and its launch shims can run. **Paperclip version or commit** `master` at `e0d9033`. The inspected release run built commit `837add0`. **Steps to reproduce** 1. Install `@agentclientprotocol/claude-agent-acp@0.73.0` and `opencode-ai@1.18.34` with pnpm 9.15.4 on a glibc Linux host. 2. List `node_modules/.pnpm`. 3. See the `*-musl` entries. Their `package.json` declares `"libc": ["musl"]`. ## What Changed - `provider-pack-runtime-tree.mjs`: add `providerPackLibc()`. It reports the libc of the Node that the pack copies (`glibc` or `musl` on Linux, none on other platforms). - `provider-pack-runtime-tree.mjs`: add `removeForeignLibcPackagesFromProviderPack()`. It removes each pnpm virtual-store entry whose own `package.json` declares a `libc` list without the pack libc. Packages without a `libc` field stay (for example `@openai/codex-linux-*`, which holds Codex's static musl binary). - `provider-pack-runtime-tree.mjs`: add `removePnpmStoreEntries()`. It removes store entries and every symlink that resolves into them, so no dangling link stays in the pack. It refuses names outside `node_modules/.pnpm`. - `provider-pack-executable-shims.mjs`: when `provider-assets/copilot/<platform>/copilot` has the same bytes as the npm Copilot executable, the `copilot-<platform>` command now points at the asset and the npm copy is removed. If the bytes are different, the old behavior stays. - `build-provider-pack.mjs`: run the libc prune immediately after `pnpm deploy`. Log the removed entries. - No file bound by `provider-pack.json` changes. Thus the manifest digest and the qualification and digest pins do not change. ## Verification - `node --test packages/paperclip-runner/scripts/provider-pack-runtime-tree.test.mjs packages/paperclip-runner/scripts/provider-pack-executable-shims.test.mjs packages/paperclip-runner/scripts/candidate-provider-pack.test.mjs packages/paperclip-runner/scripts/release-provider-pack.test.mjs`: 40/40 pass. - The new Copilot test fails with the old `provider-pack-executable-shims.mjs` and passes with the new file. - A real pnpm 9.15.4 tree with the pack's provider dependencies, on linux-x64 (glibc): - The prune removed `@anthropic-ai+claude-agent-sdk-linux-x64-musl`, `opencode-linux-x64-musl`, and `opencode-linux-x64-baseline-musl`. The Copilot dedupe removed the npm Copilot entry. - `node_modules` changed from 1.9 GB to 1.1 GB. - `find node_modules -xtype l` found no dangling links. - The claude-agent-acp SDK still resolves `@anthropic-ai/claude-agent-sdk-linux-x64/claude`. - `opencode --version` prints `1.18.34`. - The rewritten `copilot-linux-x64` prints `GitHub Copilot CLI 1.0.88.` - Earlier measurement: we pruned the same paths from the extracted pack artifacts and compressed them again with `tar -czf`. linux-x64 went from 1131 MB to 876 MB. linux-arm64 went from 519 MB to 360 MB. - Not run locally: a full `build-provider-pack.mjs` build. This sandbox has 10 GB of disk, which is not sufficient for a full workspace install and a 3 GB pack. The `release-verify.yml` provider-pack step and the Daytona image `RUN claude/codex/acpx --version` checks run the full build. - `test/acpx-codex-package-contract.test.mjs` has 1 failure in this sandbox, and the failure is the same on `master`. The cause is that this checkout has no `node_modules`. ## Risks - Low risk. All pack builds run on glibc hosts (Debian bookworm images and Ubuntu CI). On a musl build host, the prune removes the glibc packages, and the existing Claude executable resolution fails at build time. Before this change, that build gave a pack with a Claude binary that could not run. - The `copilot-linux-x64` command now runs the binary from `provider-assets/copilot/linux-x64/`. The bytes are identical. Runtime admission already uses this asset, through the descriptor-pinned lease. - The pack content changes but the manifest digest does not. The remote verifier checks only the bound artifacts, `dist`, and provider asset trees. None of these change. - Not in this pull request: Pi's bundled Node is byte-identical to the pack Node (121 MB, about 43 MB compressed). A fix for it changes the qualified Pi closure digest. Also, the `release.yml` publish jobs download the full packs, but they read only `provider-pack.json`. These are possible follow-ups. ## Model Used - Claude Opus 5.5 (Anthropic), model ID `claude-opus-5-5`, used through Claude Code with tool use (shell, file edit) and code execution in a sandbox. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes (no doc change is necessary: build behavior and commands do not change) - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [ ] I will address all Greptile and reviewer comments before requesting merge 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: devinfoley <139239+devinfoley@users.noreply.github.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
Devin FoleyandPaperclip
|
b5bf86ea09 |
Clear process identity when settled native runs stop (#17374)
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work
> - Each heartbeat run stores the process identity of its runner in
`heartbeat_runs.process_pid`, `process_group_id` and
`process_started_at`
> - Only cancellations (`release_run_resources`), restart recovery and
heartbeat recovery cleared these fields
> - Failed native runs take the `schedule_retry` path, so they kept a
PID after the run finished, for weeks
> - Liveness checks for old runs then depend on PID-reuse defenses, and
code that reads a set PID as "running" is misled
> - This pull request releases the process identity of settled native
runs when the process is confirmed gone, and backfills old sandbox rows
> - The benefit is that a finished run no longer looks like it owns a
live process, while live, warm and retrying processes keep their
identity
## Linked Issues or Issue Description
Refs #17358. Bug report described in-PR (bug report template).
**What happened?**
On one instance, 16 of 23 old native workspace sync folders belonged to
failed or cancelled runs that finished more than a week earlier and
still had `process_pid` set. The PID is written when the run's process
starts (for remote targets it is the sandbox PID). The status arbiter
clears it only through `release_run_resources`, which only cancellations
emit. Failed native runs take `schedule_retry`, and other terminal paths
(finalization reconciler, restart recovery terminal failures, watchdog
timeouts) do not clear it either.
**Expected behavior**
When a run is terminal and its process is confirmed gone (or no longer
owned), the process identity is cleared. A process that may still be
alive keeps its identity.
**Steps to reproduce**
1. Run a native task whose provider turn fails, so the arbiter decides
`run_failed_partial_evidence_preserved` with a `schedule_retry` effect.
2. Wait for the run to finish and the runner to exit.
3. Read the run row: `status = 'failed'` and `process_pid` /
`process_group_id` / `process_started_at` are still set.
**Paperclip version or commit:** `master` at
|
||
|
Nicky LeachandPaperclip
|
a3e8fa568a |
Reduce repeated work during plugin Runner admission (#17377)
## Thinking Path > - Paperclip manages agents and their execution authority. > - A plugin can host Runner tasks on a persistent execution host. > - Admission checks must observe current lifecycle state and credentials. > - The current runtime reader also resolves files, skills, and connections for each check. > - Some providers also repeat the same project authentication check during submit. > - This pull request adds a compact fresh reader and an explicit submit-time validation contract. > - Providers can remove repeated work while Core keeps lease and scope checks. ## Linked Issues or Issue Description **What existing behavior does this improve?** Plugin Runner startup before task submission. **Current behavior** One measured startup made 30 full runtime reads during bootstrap. These reads took 5.759 seconds. Each guard can resolve instructions, skills, and connection assignments even when it only needs current identity and lifecycle state. Providers that already enforce project authentication in submit must also perform a separate Core preflight. **Proposed behavior** Use a lifecycle-owned runtime identity reader for guards. Keep the full runtime reader for synchronization inputs. A trusted provider can declare `projectAuthCheckAtSubmit: true` in its prepared result. Core pins the exact configuration, provider, and project set on the lease. The provider must reject authentication conflicts before credential installation or task start. Providers that omit the flag retain the existing preflight. **Reason and benefit** Estimated savings with a paired consumer: 4–5 seconds from compact runtime checks, plus roughly 0.5–0.9 seconds from removing the duplicate authentication operation. These estimates overlap and are not a measured deployed result. Refs #17367 for the separate run-log HEAD optimization; this PR does not duplicate it. **Breaking changes** No migration. The SDK adds a host method and an optional prepared-result field. Consumers that use them require this Core host version. Older providers retain their existing flow. ## What Changed - Add `agents.getRuntimeIdentity` across SDK, RPC, host service, and test fixtures. - Reuse the existing company, plugin ownership, readiness, and capability checks. Read lifecycle state and protected instance secrets fresh on every call. - Skip instruction, organization, skill, and connection hydration in the compact reader. - Allow explicit submit-time authentication enforcement, with prepared scope pinned and checked again at dispatch. - Emit fixed preparation, project-auth, submit, and connection timing phases. ## Verification - 72 targeted SDK capability, plugin task-owner, and environment-task mediator tests passed. - 33 real embedded-Postgres lifecycle and secret tests passed, including pause, content revision, company isolation, plugin disablement, and secret rotation. - SDK and server TypeScript checks passed. Local Runner/eval TypeScript outputs were built to satisfy server type resolution. - Local Postgres required missing library aliases in ignored dependencies; no product source was changed for that repair. - Full repository suites are delegated to GitHub CI. No live rollout or speedup claim. ## Risks The submit-time flag delegates enforcement to a trusted provider. It is not proof that prepare already checked authentication. The provider must check before VM credential installation or task admission. Core still checks exact prepared scope, active lease, run identity, credential type, and dispatch fencing. Existing providers keep the separate check. Roll back a paired consumer before rolling back this host. ## Model Used OpenAI GPT-6 (Codex), with code execution and repository tools. Exact deployment build and context-window size are not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links - [x] My branch name describes the change and contains no internal ticket id - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Nicky Leach
|
2e38649341 | perf(run-logs): skip durable log restore for a proven-new run (#17367) |